eXPRESS Polymer Letters Vol.14, No.

8 (2020) 768–779
Available online at www.expresspolymlett.com
https://doi.org/10.3144/expresspolymlett.2020.63

Processing induced segregation in PLA/TPS blends: Factors

and consequences
M. Józó1,2, L. Cui1,2, K. Bocz3, B. Pukánszky1,2*
1
Laboratory of Plastics and Rubber Technology, Department of Physical Chemistry and Materials Science, Budapest
University of Technology and Economics, H-1521 Budapest, P.O. Box 91, Hungary
2
Institute of Materials and Environmental Chemistry, Chemical Research Center, H-1519 Budapest, P.O. Box 286,
Hungary
3Department of Organic Chemistry and Technology, Budapest University of Technology and Economics, H-1521
Budapest, P.O. Box 91, Hungary

Received 6 November 2019; accepted in revised form 21 January 2020

Abstract. Poly(lactic acid) (PLA) and thermoplastic starch (TPS) blends with two different glycerol contents were prepared
by injection molding. Mechanical properties were characterized by tensile and impact testing, structure by scanning electron
microscopy (SEM), Fourier transform infrared (FTIR) as well as Raman spectroscopy, and water absorption was determined
as a function of time. Compression-molded specimens were used as reference. The properties of the blends cover a wide
range, stiffness changes from 3.3 to around 1.0 GPa, while strength from 54 to 22 MPa as TPS content increases from 0 to
50 wt%. Heterogeneous structure forms in the blends because of the weak interaction of the components. Processing condi-
tions do not change bulk properties. Weak interactions and the large difference in the viscosity of the components lead to the
formation of a skin on the surface of the specimens. The skin consists mainly of PLA, while the core contains a larger amount
of TPS. The thickness of the skin depends on processing technology and conditions; it is about 18 µm for the injection-
molded, while 4.5 µm for the compression-molded parts at 50 wt% TPS content. The development of the skin layer can be
advantageous in some applications because it slows down water absorption considerably.

Keywords: biopolymers, skin and core structure, interactions, mechanical properties, water absorption

1. Introduction fuel resources, the problems of the accumulation of

In the last few decades, biopolymers came into the plastic waste and offer a neutral carbon footprint,
focus of attention. Considerable research is done on which is becoming more and more important aspect
these materials, and their use in everyday practice of environmental protection. Moreover, abundant and
increases continuously as well. Various numbers are renewing resources are available from natural poly-
published about their growth rate from 6 to 15% [1], mers, like cellulose, lignin, natural fibers, chitin, etc.
but even the smallest is larger than the growth rate and they are usually also cheap at the same time.
of GDP in Europe; biopolymers are obviously the Besides their advantages, biopolymers also have some
polymeric materials of the future. This significant drawbacks. They are often sensitive to water and
interest and growth rate have many reasons. A con- heat, they are difficult to process with the traditional
siderable number of biopolymers are based on natu- technologies of plastic processing and the properties
ral resources while others are biodegradable. Accord- of the products are usually inferior to those prepared
ingly, they answer the questions of depleting fossil from commodity polymers. Biopolymers are often

*
Corresponding author, e-mail: [email protected]
© BME-PT

768
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

modified by a number of ways to overcome these conditions of an industrial processing technology, in-
drawbacks. Blending is an obvious way to combine jection molding, on the morphology and properties of
the advantageous properties of various biopolymers the blends was studied in detail. The possible devel-
and compensate for their weaknesses. Many papers opment of special structures during processing was
have been published in the literature on the combi- also checked and their effect on blend properties was
nation of PLA and wood [2–6], natural fibers [7–11] determined. Specimens prepared by compression
or nanocellulose [12–15], on aliphatic polyesters and molding earlier [35] were used as a reference in the
starch [16–20], and on other combination of materi- study, and the consequences of the results for practice
als [21–26]. are also mentioned in the last section of the paper.
Both PLA and starch are relatively cheap and they
are available in large quantities. However, the direct 2. Experimental
combination of starch powder and PLA results in an 2.1. Materials
inhomogeneous material with rather poor properties The PLA used was the Ingeo 4032D grade (Mn =
[27, 28]. Plasticized, thermoplastic starch (TPS) is 88500 g/mol and Mw/Mn = 1.8) produced by Nature-
blended more often with PLA to obtain a more ho- Works (Minnetonka, MN, US), which is recom-
mogeneous material [29, 30]. However, the two com- mended for extrusion by the producer. The polymer
ponents are immiscible and dispersed structure forms (<2% D isomer) has a density of 1.24 g·cm–3, while
upon blending [31]. Opinions about the compatibil- its melt flow rate (MFR) is 3.9 g/10 min at 190 °C
ity or interaction of the components are considerably and 2.16 kg load. The corn starch used for the prepa-
divided, some authors claim good compatibility [32, ration of TPS was supplied by Hungrana Ltd., Sza-
33], while others complete immiscibility [34]. In a badegyháza, Hungary, and its water content was
previous study, we prepared two series of PLA/TPS 12 wt%. Glycerol with 0.5 wt% water content was
blends and determined their structure and properties obtained from Molar Chemicals Ltd., Halásztelek,
[35]. We showed that the glycerol used for the plas- Hungary and it was used for the plasticization of
ticization of starch stays in that phase even after starch without further purification or drying. Ther-
blending, and by using thermodynamic considera- moplastic starch samples containing 36 and 47 wt%
tions we proved that the interaction between the two glycerol (TPS36 and TPS47, respectively) were pre-
components is rather weak. Because of weak inter- pared and used in the experiments. The TPS content
facial adhesion, the properties of the blends are mod- of the injection-molded PLA/TPS blends was 0, 5,
erate. 10, 15, 20, 30, 40 and 50 wt%, while the composi-
The blends were prepared by compression molding tion of the compression-molded samples changed
in the study mentioned above [35]. We do not know from 0 to 100 wt% in 10 wt% steps.
anything about the behavior of the blends under prac-
tical processing conditions like extrusion or injection 2.2. Sample preparation, processing
molding. In heterogeneous materials, various changes Corn starch was dried in an oven before composite
may occur in the structure under such conditions as preparation (105 °C, 24 hours). The thermoplastic
the exfoliation of clays [36], the attrition of fibers starch powder was prepared by dry-blending in a
[37] or the segregation of the components [21]. Seg- Henschel FM/A10 (Zeppelin Systems GmbH., Fried-
regation was observed in polymers filled with glass richshafen, Germany) high-speed mixer at 2000 rpm.
beads [38–40], containing an elastomer impact mod- TPS was produced by processing the dry-blend on a
ifier [41], block copolymers [42, 43] or small molec- Rheomex 3/4" single screw extruder (Haake Technik
ular weight additives [44, 45]. Since PLA and TPS GmbH, Vreden, Germany) attached to a Haake
form heterogeneous blends with weak interactions Rheocord EU 10 V (Haake Technik GmbH, Vreden,
between the components, structural phenomena, first Germany) driving unit with the temperature profile
of all, segregation, may occur also in them [46, 47]. of 140–150–160–170 °C and at 60 rpm screw speed.
Taking into account all these considerations, the goal For the preparation of the compression-molded plates,
of our work was to prepare specimens from PLA/ PLA and the second component were homogenized
TPS blends and determine their structure and prop- in an internal mixer (Brabender W 50 EHT, Braben-
erties. Two series of blends were prepared using TPS der GmbH & Co. KG, Duisburg, Germany) at 190 °C
with different plasticizer contents. The effect of the and 50 rpm for 12 min. Before homogenization

769
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

poly(lactic acid) was dried in a vacuum oven Unterhaching, Germany). The spectrograph was set
(110 °C, 4 hours). Subsequently, the melt was trans- to provide a spectral range of 350–1550 cm–1 with
ferred to a Fontijne SRA 100 compression molding 2 cm–1 resolution. First, reference spectra were col-
machine (Fontijne Presses, Delft, The Netherlands) lected from the neat PLA and TPS materials with an
(190 °C, 5 min) to produce 1 mm thick plates for fur- objective providing a magnification of 50 using the
ther testing. acquisition time of 40 s and averaging 3 measured
TPS and PLA were homogenized in a Brabender spectra at each point. Subsequently, Raman spectra
DSK 42/7 twin-screw compounder (Brabender were collected along the cross-section of the blends
GmbH & Co. KG, Duisburg, Germany) before in- to create line maps and to determine the thickness of
jection molding. Set temperatures were 170–175– the PLA skin layer of the samples. All spectra were
180–180 °C and the speed of the screws was 30 rpm baseline corrected and normalized before multivari-
during extrusion. Standard 4 mm thick ISO 179 type ate evaluation in order to eliminate the intensity de-
tensile specimens were injection molded using a viation among the measured points. The spectral
Demag IntElect 50/330-10 type all-electric injection concentration of the components was determined
molding machine (Sumitomo Demag, Tokyo, Japan) point by point by classical least squares (CLS) mod-
with the temperature profile of 180–180–185–190 °C, eling using the reference spectra collected from neat
at 20 °C mold temperature, 1500 bar injection and PLA and TPS; i.e. each Raman spectrum of the blends
650 bar holding pressure (decreasing to 0 bar in taken from any location was assumed a linear com-
12 s) and 50 s cooling time. bination of the two reference spectra. The structure
of the blends was studied by scanning electron mi-
2.3. Characterization, measurements croscopy (SEM) using a Jeol JSM 6380 LA appara-
Mechanical properties were characterized by tensile tus (Jeol USA Inc., Peabody, MA, US). Samples were
testing using an Instron 5566 universal testing ma- broken at liquid nitrogen temperature and then a
chine (Illinois Tool Works Inc, Norwood, MA, US). smooth surface was created by cutting the sample
The gauge length was 115 mm; tensile modulus was with a microtome. Surfaces were etched with 1 mol
determined at 0.5 mm/min, while properties meas- HCl to remove TPS particles. Water absorption was
ured at larger deformations at 5 mm/min cross-head determined at 23 °C and 52% relative humidity by
speed. Five parallel measurements were carried out the measurement of the weight increase of the spec-
at each blend composition. Impact resistance was de- imens. The desired relative humidity was achieved
termined on notched Charpy type specimens (ISO with a saturated solution of Mg(NO3)2.
179) at 2 mm notch depth using a Ceast Resil 5.5
equipment (Illinois Tool Works Inc, Norwood, MA, 3. Results and discussion
USA) both in normal as well as in instrumented im- The results are presented in several sections. The bulk
pact testing. The viscosity of the components was properties of the blends and the evolution of their
determined by oscillatory rheometry using an Anton structure as a function of composition are discussed
Paar Physica MCR 301 rheometer (Anton Paar in the first two followed by the presentation of the
GmbH, Graz, Austria) in the plate-plate geometry segregation of the components and the considerations
from 0.1 to 600 s–1 frequency. The gap was 1 mm, about the reasons for this phenomenon in the subse-
and the temperature was maintained at 190 °C. The quent two. Consequences for properties and practice
composition was analyzed by Fourier transform in- are discussed in the last section of the paper.
frared spectroscopy (FTIR). Spectra were recorded
using a Bruker Tensor 27A (Bruker BioSpin Corpo- 3.1. Bulk properties
ration, Billerica, MA, US) apparatus with a Bruker The composition dependence of the properties of
Platinum ATR probe (Bruker BioSpin Corporation, PLA/TPS blends is determined by two main factors,
Billerica, MA, US). Spectra were recorded from the large difference in the macroscopic properties of
4000 to 400 cm–1 at 2 cm–1 resolution with 32 scans. the components and their weak interaction [35]. The
Raman spectra were collected using a Horiba Jobin- stiffness of the blends is plotted against their TPS
Yvon LabRAM system coupled with an external content in Figure 1. The effect of the first factor is
785 nm diode laser source and an Olympus BX-40 clearly seen in the figure, stiffness decreases steeply
optical microscope (Horiba Jobin Yvon GmbH, and monotonously with increasing TPS content. The

770
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

Figure 1. Young’s modulus of PLA/TPS blends plotted as a Figure 3. Dependence of the notched Charpy impact strength
function of starch (TPS) content. Effect of the de- of PLA/TPS blends on composition.
gree of plasticization. Symbols: (○) TPS36, (□) TPS47.
Symbols: (○) TPS36, (□) TPS47.
differences can be observed in the strength of the
effect of the second factor does not appear since blends prepared with the two types of TPS materials,
stiffness is influenced only slightly by interfacial ad- TPS36 and TPS47. The reason for this small difference
hesion [48]. Plasticization, on the other hand, influ- is the heterogeneous structure of the blends and the
ences the inherent properties of TPS, thus Young’s weak interaction of the components. Somewhat sur-
modulus of the blends decreases with increasing prisingly, the composition dependence of deformabil-
plasticizer content. ity is different for the two series of blends; elongation-
Properties measured at larger deformations, i.e. ten- at-break decreases continuously for the PLA/TPS36
sile strength and elongation-at-break in this case, blends but exhibits a maximum for the other series.
show a somewhat different picture (Figure 2). Ten- Obviously, the softer particles of TPS47 and its partial
sile strength decreases continuously with increasing solubility in the PLA matrix [35] results in the increase
TPS content, similarly to modulus, and only slight of deformability at intermediate TPS contents.
The dissimilar deformability of the two series of
blends results in different impact resistance as well.
As Figure 3 shows, the impact strength of the PLA/
TPS36 blends decreases continuously with increas-
ing TPS content, while that of the PLA/TPS47 blends
increases slightly above a certain TPS content, above
20 vol%. This increase is caused by the same factors
mentioned above, and it could be regarded benefi-
cial, apart from the fact that the absolute values of
impact resistance are rather small; they do not ex-
ceed 3 kJ/m2 that is not sufficient for several practi-
cal applications. We can conclude from these results
that the bulk properties of the studied blends are gov-
erned by component properties, structure and the
Figure 2. Effect of TPS content and the extent of plasticiza- weak interaction of the components.
tion on the ultimate tensile properties of PLA/TPS
blends. 3.2. Structure
Symbols: (○, ●) TPS36, (□, ■) TPS47, The heterogeneous structure of the blends has been
empty symbols: tensile strength, mentioned several times in the previous section, but
full symbols: elongation-at-break.
no evidence was supplied to support the statements.

771
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

However, previous research showed the incompati-

bility of the components and the development of
weak interactions; thus the assumption of the forma-
tion of a heterogeneous structure seemed to be ob-
vious. The SEM micrographs presented in Figure 4
confirm this assumption quite strongly. Voids left by
dispersed TPS particles after etching are visible on
the SEM micrographs shown in Figure 4a recorded
on the cut surface of a specimen containing 10 wt%
of the TPS47 starch. The size of the particles is rather
large, in the range of 5–10 µm, which justifies the
strong decrease of tensile strength with increasing
TPS content (see Figure 2). The size of the particles
increases somewhat with increasing TPS content as
expected (Figure 4b), and some of them touch each
other forming larger aggregates. The coalescence of
the particles depends on their interaction with the
matrix polymer and on their viscosity. The first fac-
tor facilitates, while the second hinders coalescence
and the viscosity of TPS is rather large.
Structure develops further with increasing TPS con-
tent. The size of the particles increases even more and
they start to form an interpenetrating network (IPN)
like structure (Figure 4c). The formation of such a
structure cannot be confirmed based on the presented
and similar micrographs but must occur at some com-
position, since TPS becomes the continuous phase
at large TPS content [35]. However, the composition
range of the IPN structure is very narrow because of
the weak interaction and poor compatibility of the
components. We must also conclude that the maxi-
mum in deformability in Figure 2 is not the result of
the formation of an IPN structure or phase inversion
since this later occurs at larger concentration.

3.3. Segregation, skin formation

The micrographs presented in Figure 4 taken from
compression-molded plates offered a rather uniform
picture about the structure of the blends. Apart from
the association of dispersed TPS particles and a hint
about the formation of an IPN structure, no other Figure 4. SEM micrographs recorded on PLA/TPS47 blends.
structural phenomenon can be observed on them. Effect of starch content. Compression-molded
specimens. TPS content: a) 10 wt%, b) 30 wt%,
Micrographs recorded on injection molded speci- c) 50 wt%.
mens offer a somewhat different picture as shown
by Figure 5. First of all, dispersed TPS particles are recorded at different magnifications and both show
elongated, they are oriented parallel to the wall of convincingly the development of a skin layer. More-
the mold. The micrographs also include the surface over, Figure 5a indicates that the skin surrounds the
of the specimen and indicate the formation of a layer entire specimen. The micrographs taken from the
with a different composition from that of the core. compression molded parts (Figure 4) indicate the ho-
The two micrographs presented in Figure 5 were mogeneous distribution of TPS within the specimen,

772
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

Figure 5. Structure of injection-molded PLA/TPS47 specimens. The orientation of TPS droplets and the development of a
skin layer. TPS content is 50 wt%. Magnification: a) 400×, b) 1600×.

and the development of a skin layer cannot be sus- practically at all, while the skin contains mainly PLA
pected at all, at least based on those micrographs. with a small amount of dispersed TPS. The presence
However, closer scrutiny and investigation showed of TPS is proved by the peak appearing at 1020 cm–1
that a skin layer forms also around the compression and the broad peak around 3300 cm–1 both assigned
molded plates, only its thickness is somewhat small- to the –OH groups of starch and glycerol. The almost
er than in the case of injection-molded specimens. complete lack or a very small amount of PLA in the
The thickness of the skin formed is around 18 µm in core is quite surprising since the micrographs in Fig-
injection and approximately 4.5 µm in compression ure 5 clearly show the continuous PLA phase, which
molding. contains dispersed TPS particles. The extinction co-
FTIR spectra were recorded on the skin and the core efficient of the ester group appearing at 1745 cm–1
in order to identify the main components and to de- is rather large, thus even small amounts can be de-
termine their composition. The spectra are presented tected with infrared spectroscopy. However, the band
in Figure 6, together with that of the neat PLA and of the ester group is extremely small showing the ab-
TPS as reference. The comparison of the spectra sence or very small concentration of PLA in the core.
clearly shows that the core does not contain PLA The only explanation we can find is that the tech-
nique used, i.e. attenuated total reflection spec-
troscopy, scans only a smaller area [49] in which the
TPS component dominates.
Raman spectroscopy was also used in order to fur-
ther explore the phenomenon of skin formation and
characterize the composition of the layer. The spec-
tra were taken along a line from the surface towards
the core. The composition of the core and the skin is
presented in Figure 7 as a function of position meas-
ured from the surface. The results confirm previous
observations and indicate that the skin consists main-
Figure 6. FTIR spectra recorded on the skin and core of in- ly of PLA, while the amount of TPS is larger in the
jection-molded PLA/TPS47 specimens containing core. The exact concentration of the two structural
50 wt% starch. The spectra of neat PLA and TPS formations is difficult to determine because of the
are included as a reference. The spectra were limitations of the measurement techniques used, as
recorded in the ATR mode.
mentioned above. The explanation for the segregation

773
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

poly(ethylene terephthalate), to improve its wetta-

bility and printability. Numerous factors determine
the migration of one component during flow includ-
ing surface tension [59, 60], interfacial tension to-
wards the surface of the mold [51], the interaction
of the components [18], molecular weight [61], mo-
bility that is related to viscosity [62, 63] and deter-
mines the rate of migration.
The migration of a component during flow has been
studied extensively, mostly in dilute polymer solu-
tions. Dill and Zimm [64] developed the Equa-
tion (1) for the description of the rate of migration:
2
h m co 5
v = rk T R (1)
B

Figure 7. The composition of PLA/TPS47 blends as a func- where v is the migration velocity in the radial direc-
tion of position measured from the surface of the
tion, r is the radial position, R is the chain end dis-
specimen. The composition was determined by
Raman spectroscopy.
tance of the polymer chain in equilibrium depending
on molecular weight, ηm is the viscosity of the ma-
of the components and the formation of the skin and trix polymer, T is the temperature, kB is Boltzman’s
core structure needs further considerations. constant and γ· is the shear rate. The equation consid-
ers the effect of molecular weight, the viscosity of
3.4. Considerations, discussion the matrix, and shear conditions but does not take
Segregation in polymeric materials containing more into account interactions and the relative viscosity
than one component has been observed many times. of the components although these seem to be essen-
The typical phenomenon is the plate out of additives tial factors in segregation [64, 65]. Khan et al. [66]
during processing or the separation of two polymers assumed that the inhomogeneous stress field is the
during the cleaning of an injection molding machine. driving force, which results in migration towards
Segregation has been observed in two directions, to- smaller stresses. They assumed strong interaction
wards the center of the flow or towards the edge of among polymer molecules and thus developed Equa-
the part. Segregation towards the center was observed, tion (2) for the rate of migration:
for example, by Szalánczi and Kubát [50], who injec-
0.0085nL5
tion molded polymers containing large glass beads to v= (2)
kTco 2 r
study the composition along the flow path and across
it. They observed the accumulation of the beads to- where ν is the characteristic migration velocity, μ is
ward the end of the part. Similarly, Karger-Kocsis and the viscosity of the fluid, L is the half-length of the
Kiss [51] observed increased elastomer content in the particle, r is the radial coordinate, and γ· is the shear
center of injection-molded specimens and concluded rate [66]. Although the authors assumed strong in-
that the large difference in the viscosity of the com- teractions, it does not appear in Equation (2) and
ponents results in larger extent of segregation. The mi- they do not consider the possible effect of the vis-
gration of larger molecular weight components to- cosity ratio of the two components.
wards the center was explained with an entropic In our case, several factors must play a role in the
driving force, with the larger orientation of longer segregation of the components. The interaction be-
molecules close to the wall of the mold [52, 53]. tween the two components is weak; their miscibility
Migration towards the edge also occurs quite fre- is very limited [35]. The solubility parameters of the
quently. The segregation of small molecular weight three components, PLA, starch and glycerol are 24.1,
components [54], additives [55, 56], block copoly- 26.7 and 28.9 J1/2/cm3/2, which indicates considerably
mers [51] and immiscible polymers [57] was also larger polarity for TPS than for PLA. In accordance
observed many times. Rezaei Kolahchi [58] used the with the different polarity of the components, surface
phenomenon to modify the surface characteristics of tensions are also different, 48.7 and 60.3 mJ/m2 for

774
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

Figure 8. Frequency dependence of the complex viscosity Figure 9. Comparison of the bulk properties of blend spec-
of PLA and TPS47 determined by oscillatory imens prepared by compression and injection
rheometry in the plate-plate geometry. molding, respectively, from PLA/TPS47 blends.
Symbols: (○, ●) tensile strength,
PLA and TPS47, respectively, favoring the migration (□, ■) elongation-at-break,
of PLA towards the surface. The molecular weight of empty symbols: compression-molded,
starch is much larger than that of PLA and also their full symbols: injection molded.
viscosity differs considerably as shown by Figure 8,
at least at small shear rates. Because of smaller mo- injection-molded specimen than for the compression-
lecular weight and viscosity, the mobility of PLA mol- molded one. The difference evidently does not result
ecules is larger and incompatibility, as well as smaller from the thickness of the skin, but from the orienta-
surface tension, all drive the matrix polymer towards tion of the components resulting from the mold fill-
the surface, thus forming the skin layer. Processing ing process (see Figure 5). We can conclude from
conditions including different shear rates and cooling these, and from other results not shown, that bulk
conditions, must also contribute and result in the dif- properties are not influenced much by segregation
ferent thickness of the skin formed in the two process- and skin formation much.
ing technologies, i.e. compression and injection mold- Starch, and TPS even more, is very sensitive to
ing. The formation of a skin layer and its main reasons water. They absorb a considerable amount of water,
are established, the only remaining question is the ef- which modifies, usually deteriorates their properties.
fect of this structure on blend properties. One of the main benefits of blending is the decrease
in water uptake. The relative weight increase, i.e.
3.5. Consequences water uptake, of compression and injection molded
In order to assess the effect of the formation of the blend samples are compared in Figure 10. The figure
skin layer in PLA/TPS blends, first, the properties clearly indicates and calculations proved that equi-
of compression and injection molded specimens are librium water uptake is the same for the two types
compared to each other. Although compression of blends, but the rate of absorption differs consid-
molded parts also develop a skin layer, its thickness erably. Equilibrium water uptake is determined by
is much smaller than for injection-molded speci- the composition of the blends, by the amount of TPS,
mens. The tensile strength and deformability of the while the rate of absorption is considerably influ-
two kinds of samples are compared to each other in enced by the presence of the skin. Water uptake was
Figure 9. Processing technology, thickness (com- modeled to express these relationships quantitative-
pression-molded: 1 mm, injection-molded: 4 mm) ly. The following form of Fick's law was fitted to the
and skin layer do not have any effect on strength. experimental points to determine the overall rate of
Deformability, on the other hand, is larger for the absorption (see Equation (3)):
8 # Q- V + 1 1
Mt = M3 G1 - 9 exp Q- 9at V + 25 exp Q- 25at V
exp at &J (3)
r2

775
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

parts than for their compression-molded counter-

parts. The result has considerable practical conse-
quences. Slower water uptake changes the properties
of the product and modifies its lifetime. In the case
of products with a determined or limited lifetime,
this difference can be of practical relevance.

4. Conclusions
The study of PLA/TPS blend samples prepared by in-
jection molding showed that their properties cover a
wide range depending on composition. The stiffness
of the blends changed from 3.3 to around 1.0 GPa and
their strength from 54 to 22 MPa as TPS content in-
creased from 0 to 50 wt%. The blends have heteroge-
neous structures because of the weak interaction of
Figure 10. Water absorption isotherms of PLA/TPS47 blends the components, and phase inversion cannot be ob-
of various compositions. Effect of starch (TPS) served in the studied composition range. Processing
content. conditions do not change bulk properties, the mechan-
Symbols: (○, ●) 10 wt%, (□, ■) 30 wt%,
ical properties of compression and injection molded
(, ▲) 50 wt% TPS content;
open symbols: compression molded, parts were very similar. Weak interactions and the
full symbols: injection molded. large difference in the viscosity of the components
leads to segregation, to the formation of a skin layer
where Mt and M∞ are the amount of absorbed water on the surface of the specimens. The skin consists of
at time t and at infinite time, respectively, and a is the mainly PLA, while the core contains a larger amount
overall rate of absorption. The determined rates are of TPS. The thickness of the skin depends on process-
plotted against TPS content in Figure 11. The rate of ing technology and conditions; it is about 18 µm for
water uptake is much smaller for the injection molded the injection-molded, while 4.5 µm for the compres-
sion-molded parts at 50 wt% TPS content. The devel-
opment of the skin layer can be advantageous in some
applications because it slows down water absorption
considerably. Bulk properties must be improved for
practical applications by the modification of interac-
tions, by coupling, for example.

Acknowledgements
The significant help of Péter Müller, József Bere, and Erika
Fekete Bódiné in sample preparation and measurements is
highly appreciated. The authors acknowledge the financial
support of the National Scientific Research Fund of Hungary
(OTKA Grant No. K 120039 and FK 129270) for this project
on the modification of polymeric materials.

References
[1] Dawande R.: Bioplastics market by type (biodegradable
Figure 11. Effect of TPS content and processing technology
plastic and non-biodegradable plastic) and application
on the rate of water absorption in PLA/TPS47
(rigid packaging, flexible packaging, textile, agriculture
blends.
& horticulture, consumer good, automotive, electronic,
Symbols: (□) compression molded,
building & construction, and others) – Global opportu-
(■) injection molded.
nity analysis and industry forecast, 2018–2024. Allied
Market Research, Portland (2017).

776
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

[2] Csikós Á., Faludi G., Domján A., Renner K., Móczó J., [13] Kian L. K., Saba N., Jawaid M., Sultan M. T. H.: A re-
Pukánszky B.: Modification of interfacial adhesion with view on processing techniques of bast fibers nanocel-
a functionalized polymer in PLA/wood composites. Eu- lulose and its polylactic acid (PLA) nanocomposites.
ropean Polymer Journal, 68, 592–600 (2015). International Journal of Biological Macromolecules,
https://doi.org/10.1016/j.eurpolymj.2015.03.032 121, 1314–1328 (2019).
[3] Faludi G., Dóra G., Renner K., Móczó J., Pukánszky B.: https://doi.org/10.1016/j.ijbiomac.2018.09.040
Improving interfacial adhesion in PLA/wood biocom- [14] Zhang Y., Cui L., Xu H., Feng X., Wang B., Pukánszky
posites. Composites Science and Technology, 89, 77–82 B., Mao Z., Sui X.: Poly(lactic acid)/cellulose nanocrys-
(2013). tal composites via the Pickering emulsion approach:
https://doi.org/10.1016/j.compscitech.2013.09.009 Rheological, thermal and mechanical properties. Inter-
[4] Csizmadia R., Faludi G., Renner K., Móczó J., Pukánszky national Journal of Biological Macromolecules, 137,
B.: PLA/wood biocomposites: Improving composite 197–204 (2019).
strength by the chemical treatment of the fibers. Com- https://doi.org/10.1016/j.ijbiomac.2019.06.204
posites Part A: Applied Science and Manufacturing, 53, [15] Hegyesi N., Zhang Y., Kohári A., Polyák P., Sui X.,
46–53 (2013). Pukánszky B.: Enzymatic degradation of PLA/cellulose
https://doi.org/10.1016/j.compositesa.2013.06.003 nanocrystal composites. Industrial Crops and Products,
[5] Ecker J. V., Haider A., Burzic I., Huber A., Eder G., 141, 111799/1-111799/8 (2019).
Hild S.: Mechanical properties and water absorption be- https://doi.org/10.1016/j.indcrop.2019.111799
haviour of PLA and PLA/wood composites prepared by [16] Soares F. C., Yamashita F., Müller C. M. O., Pires A. T.
3D printing and injection moulding. Rapid Prototyping N.: Effect of cooling and coating on thermoplastic
Journal, 25, 672–678 (2019). starch/poly(lactic acid) blend sheets. Polymer Testing,
https://doi.org/10.1108/RPJ-06-2018-0149 33, 34–39 (2014).
[6] Dong Y., Milentis J., Pramanik A.: Additive manufac- https://doi.org/10.1016/j.polymertesting.2013.11.001
turing of mechanical testing samples based on virgin [17] Li H., Huneault M. A.: Comparison of sorbitol and glyc-
poly (lactic acid) (PLA) and PLA/wood fibre compos- erol as plasticizers for thermoplastic starch in TPS/PLA
ites. Advances in Manufacturing, 6, 71–82 (2018). blends. Journal of Applied Polymer Science, 119, 2439–
https://doi.org/10.1007/s40436-018-0211-3 2448 (2011).
[7] Mazzanti V., Pariante R., Bonanno A., de Ballesteros https://doi.org/10.1002/app.32956
O. R., Mollica F., Filippone G.: Reinforcing mecha- [18] Koh J. J., Zhang X., Kong J., He C.: Compatibilization
nisms of natural fibers in green composites: Role of of multicomponent composites through a transitioning
fibers morphology in a PLA/hemp model system. Com- phase: Interfacial tensions considerations. Composites
posites Science and Technology, 180, 51–59 (2019). Science and Technology, 164, 34–43 (2018).
https://doi.org/10.1016/j.compscitech.2019.05.015 https://doi.org/10.1016/j.compscitech.2018.05.030
[8] Gunti R., Ratna Prasad A. V., Gupta A. V. S. S. K. S.: [19] Yokesahachart C., Yoksan R.: Effect of amphiphilic
Mechanical and degradation properties of natural fiber- molecules on characteristics and tensile properties of
reinforced PLA composites: Jute, sisal, and elephant thermoplastic starch and its blends with poly(lactic
grass. Polymer Composites, 39, 1125–1136 (2018). acid). Carbohydrate Polymers, 83, 22–31 (2011).
https://doi.org/10.1002/pc.24041 https://doi.org/10.1016/j.carbpol.2010.07.020
[9] Siakeng R., Jawaid M., Ariffin H., Sapuan S. M., Asim [20] Móczó J., Kun D., Fekete E.: Desiccant effect of starch
M., Saba N.: Natural fiber reinforced polylactic acid in polylactic acid composites. Express Polymer Letters,
composites: A review. Polymer Composites, 40, 446– 12, 1014–1024 (2018).
463 (2019). https://doi.org/10.3144/expresspolymlett.2018.88
https://doi.org/10.1002/pc.24747 [21] Sui G., Jing M., Zhao J., Wang K., Zhang Q., Fu Q.: A
[10] Scaffaro R., Lopresti F., Botta L.: PLA based biocom- comparison study of high shear force and compatibiliz-
posites reinforced with Posidonia oceanica leaves. Com- er on the phase morphologies and properties of poly-
posites Part B: Engineering, 139, 1–11 (2018). propylene/polylactide (PP/PLA) blends. Polymer, 154,
https://doi.org/10.1016/j.compositesb.2017.11.048 119–127 (2018).
[11] Li X., Hegyesi N., Zhang Y., Mao Z., Feng X., Wang B., https://doi.org/10.1016/j.polymer.2018.09.005
Pukánszky B., Sui X.: Poly(lactic acid)/lignin blends [22] Arrieta M. P., Fortunati E., Dominici F., López J., Kenny
prepared with the Pickering emulsion template method. J. M.: Bionanocomposite films based on plasticized
European Polymer Journal, 110, 378–384 (2019). PLA–PHB/cellulose nanocrystal blends. Carbohydrate
https://doi.org/10.1016/j.eurpolymj.2018.12.001 Polymers, 121, 265–275 (2015).
[12] Ghasemi S., Behrooz R., Ghasemi I., Yassar R. S., Long https://doi.org/10.1016/j.carbpol.2014.12.056
F.: Development of nanocellulose-reinforced PLA nano-
composite by using maleated PLA (PLA-g-MA). Jour-
nal of Thermoplastic Composite Materials, 31, 1090–
1101 (2018).
https://doi.org/10.1177/0892705717734600

777
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

[23] Zembouai I., Kaci M., Bruzaud S., Dumazert L., [34] Mittal V., Akhtar T., Luckachan G., Matsko N.: PLA,
Bourmaud A., Mahlous M., Lopez-Cuesta J. M., Grohens TPS and PCL binary and ternary blends: Structural char-
Y.: Gamma irradiation effects on morphology and prop- acterization and time-dependent morphological changes.
erties of PHBV/PLA blends in presence of compatibi- Colloid and Polymer Science, 293, 573–585 (2015).
lizer and Cloisite 30B. Polymer Testing, 49, 29–37 https://doi.org/10.1007/s00396-014-3458-7
(2016). [35] Müller P., Bere J., Fekete E., Móczó J., Nagy B., Kállay
https://doi.org/10.1016/j.polymertesting.2015.11.003 M., Gyarmati B., Pukánszky B.: Interactions, structure
[24] Arruda L. C., Magaton M., Bretas R. E. S., Ueki M. M.: and properties in PLA/plasticized starch blends. Poly-
Influence of chain extender on mechanical, thermal and mer, 103, 9–18 (2016).
morphological properties of blown films of PLA/PBAT https://doi.org/10.1016/j.polymer.2016.09.031
blends. Polymer Testing, 43, 27–37 (2015). [36] Rajesh J. J., Soulestin J., Lacrampe M. F., Krawczak P.:
https://doi.org/10.1016/j.polymertesting.2015.02.005 Effect of injection molding parameters on nanofillers
[25] de Lucas-Freile A., Sancho-Querol S., Yánez-Pacios A. dispersion in masterbatch based PP-clay nanocompos-
J., Marín-Perales L., Martín-Martínez J. M.: Blends of ites. Express Polymer Letters, 6, 237–243 (2012).
ethylene-co-vinyl acetate and poly(3-hydroxybutyrate) https://doi.org/10.3144/expresspolymlett.2012.26
with adhesion property. Express Polymer Letters, 12, [37] Bailey R., Kraft H.: A study of fibre attrition in the pro-
600–615 (2018). cessing of long fibre reinforced thermoplastics. Inter-
https://doi.org/10.3144/expresspolymlett.2018.51 national Polymer Processing, 2, 94–101 (1987).
[26] Cailloux J., Abt T., García-Masabet V., Santana O., https://doi.org/10.3139/217.870094
Sánchez-Soto M., Carrasco F., Maspoch M. L.: Effect [38] Molina-Boisseau S., le Bolay N.: The mixing of a poly-
of the viscosity ratio on the PLA/PA10.10 bioblends meric powder and the grinding medium in a shaker bead
morphology and mechanical properties. Express Poly- mill. Powder Technology, 123, 212–220 (2002).
mer Letters, 12, 569–582 (2018). https://doi.org/10.1016/S0032-5910(01)00460-0
https://doi.org/10.3144/expresspolymlett.2018.47 [39] Kovács J. G.: Shrinkage alteration induced by segrega-
[27] Jun C. L.: Reactive blending of biodegradable poly- tion of glass beads in injection molded PA6: Experi-
mers: PLA and starch. Journal of Polymers and Envi- mental analysis and modeling. Polymer Engineering
ronment, 8, 33–37 (2000). and Science, 51, 2517–2525 (2011).
https://doi.org/10.1023/A:1010172112118 https://doi.org/10.1002/pen.22025
[28] Jang W. Y., Shin B. Y., Lee T. J., Narayan R.: Thermal [40] Balke S. T., Hu J., Joseph S., Karami A., Salerni R.,
properties and morphology of biodegradable PLA/starch Planeta M., Suhay J., Tamber H.: Polymer and particle
compatibilized blends. Journal of Industrial and Engi- separation during extrusion. in ‘Proceeding of Annual
neering Chemistry, 13, 457–464 (2007). Technical Conference for Plastics Professionals. At-
[29] Akrami M., Ghasemi I., Azizi H., Karrabi M., lanta, USA’, Vol 1, 205–211 (1998).
Seyedabadi M.: A new approach in compatibilization [41] Yamashita T., Nabeshima Y.: A study of the microscopic
of the poly(lactic acid)/thermoplastic starch (PLA/TPS) plastic deformation process in poly(methylmethacry-
blends. Carbohydrate Polymers, 144, 254–262 (2016). late)/acrylic impact modifier compounds by means of
https://doi.org/10.1016/j.carbpol.2016.02.035 small angle X-ray scattering. Polymer, 41, 6067–6079
[30] Li H., Huneault M. A.: Crystallization of PLA/thermo- (2000).
plastic starch blends. International Polymer Processing, https://doi.org/10.1016/S0032-3861(99)00856-3
23, 412–418 (2008). [42] Yan X., Liu G., Li Z.: Preparation and phase segrega-
https://doi.org/10.3139/217.2185 tion of block copolymer nanotube multiblocks. Journal
[31] Martin O., Avérous L.: Poly(lactic acid): Plasticization of the American Chemical Society, 126, 10059–10066
and properties of biodegradable multiphase systems. (2004).
Polymer, 42, 6209–6219 (2001). https://doi.org/10.1021/ja0479890
https://doi.org/10.1016/S0032-3861(01)00086-6 [43] Fukuhara K., Fujii Y., Nagashima Y., Hara M., Nagano
[32] Wang N., Yu J., Chang P. R., Ma X.: Influence of for- S., Seki T.: Liquid-crystalline polymer and block copoly-
mamide and water on the properties of thermoplastic mer domain alignment controlled by free-surface seg-
starch/poly(lactic acid) blends. Carbohydrate Polymers, regation. Angewandte Chemie, 52, 5988–5991 (2013).
71, 109–118 (2008). https://doi.org/10.1002/anie.201300560
https://doi.org/10.1016/j.carbpol.2007.05.025 [44] Bhattacharyya S. K., De S. K., Basu S.: Studies on
[33] Wang N., Yu J., Ma X.: Preparation and characterization poly(vinyl chloride)-copper composites. Part 1: State
of compatible thermoplastic dry starch/poly(lactic acid). of segregation of filler particles, electrical and mechan-
Polymer Composites, 29, 551–559 (2008). ical properties in presence of plasticizer and stabilizer.
https://doi.org/10.1002/pc.20399 Polymer Engineering and Science, 19, 533–539 (1979).
https://doi.org/10.1002/pen.760190802

778
 Józó et al. – eXPRESS Polymer Letters Vol.14, No.8 (2020) 768–779

[45] Briddick A., Li P., Hughes A., Courchay F., Martinez A., [55] Lee H., Arher L. A.: Functionalizing polymer surfaces
Thompson R. L.: Surfactant and plasticizer segregation by field-induced migration of copolymer additives. 1.
in thin poly(vinyl alcohol) films. Langmuir, 32, 864– Role of surface energy gradients. Macromolecules, 34,
872 (2016). 4572–4579 (2001).
https://doi.org/10.1021/acs.langmuir.5b03758 https://doi.org/10.1021/ma001278e
[46] Lourdin D., Coignard L., Bizot H., Colonna P.: Influ- [56] Lee H., Archer L. A.: Functionalizing polymer surfaces
ence of equilibrium relative humidity and plasticizer by surface migration of copolymer additives: Role of
concentration on the water content and glass transition additive molecular weight. Polymer, 43, 2721–2728
of starch materials. Polymer, 38, 5401–5406 (1997). (2002).
https://doi.org/10.1016/S0032-3861(97)00082-7 https://doi.org/10.1016/S0032-3861(02)00041-1
[47] Vikman M., Hulleman S. H. D., van der Zee M., [57] Qian H., Zhang Y. X., Huang S. M., Lin Z. Y.: Effect of
Myllärinen P., Feil H.: Morphology and enzymatic the surface-modifying macromolecules on the duration
degradation of thermoplastic starch–polycaprolactone of the surface functionalization. Applied Surface Sci-
blends. Journal of Applied Polymer Science, 74, 2594– ence, 253, 4659–4667 (2007).
2604 (1999). https://doi.org/10.1016/j.apsusc.2006.10.017
https://doi.org/10.1002/(SICI)1097- [58] Rezaei Kolahchi A., Ajji A., Carreau P. J.: Enhancing
4628(19991209)74:11<2594::AID-APP5>3.0.CO;2-R hydrophilicity of polyethylene terephthalate surface
[48] Gupta V. B., Mittal R. K., Sharma P. K., Mennig G., through melt blending. Polymer Engineering and Sci-
Wolters J.: Some studies on glass fiber-reinforced poly- ence, 55, 349–358 (2015).
propylene. Part II: Mechanical properties and their de- https://doi.org/10.1002/pen.23910
pendence on fiber length, interfacial adhesion, and fiber [59] You J., Liao Y., Men Y., Shi T., An L., Li X.: Composi-
dispersion. Polymer Composites, 10, 16–27 (1989). tion effect on interplay between phase separation and
https://doi.org/10.1002/pc.750100104 dewetting in PMMA/SAN blend ultrathin films. Macro-
[49] Kazarian S. G., Chan K. L. A.: ATR-FTIR spectroscop- molecules, 44, 5318–5325 (2011).
ic imaging: Recent advances and applications to bio- https://doi.org/10.1021/ma200082m
logical systems. Analyst, 138, 1940–1951 (2013). [60] Cheung Z-L., Weng L-T., Chan C-M., Hou W. M., Li L.:
https://doi.org/10.1039/C3AN36865C Morphology-driven surface segregation in a blend of
[50] Kubát J., Szalánczi Á.: Polymer-glass separation in the poly(ε-caprolactone) and poly(vinyl chloride). Lang-
spiral mold test. Polymer Engineering and Science, 14, muir, 21, 7968–7970 (2005).
873–877 (1974). https://doi.org/10.1021/la050649n
https://doi.org/10.1002/pen.760141211 [61] Chen H-L., Li L-J., Lin T-L.: Formation of segregation
[51] Karger-Kocsis J., Kiss L.: Dynamic mechanical prop- morphology in crystalline/amorphous polymer blends:
erties and morphology of polypropylene block copoly- Molecular weight effect. Macromolecules, 31, 2255–
mers and polypropylene/elastomer blends. Polymer En- (1998).
gineering and Science, 27, 254–262 (1987). https://doi.org/10.1021/ma9715740
https://doi.org/10.1002/pen.760270404 [62] Reignier J., Favis B. D.: Core–shell structure and seg-
[52] Breuer O., Tchoudakov R., Narkis M., Siegmann A.: regation effects in composite droplet polymer blends.
Segregated structures in carbon black-containing im- AlChE Journal, 49, 1014–1023 (2003).
miscible polymer blends: HIPS/LLDPE systems. Jour- https://doi.org/10.1002/aic.690490418
nal of Applied Polymer Science, 64, 1097–1106 (1997). [63] Fourati Y., Tarrés Q., Mutjé P., Boufi S.: PBAT/thermo-
https://doi.org/10.1002/(SICI)1097- plastic starch blends: Effect of compatibilizers on the
4628(19970509)64:6<1097::AID-APP9>3.0.CO;2-G rheological, mechanical and morphological properties.
[53] Miccio L. A., Liaño R., Schreiner W. H., Montemartini Carbohydrate Polymers, 199, 51–57 (2018).
P. E., Oyanguren P. A.: Partially fluorinated polymer https://doi.org/10.1016/j.carbpol.2018.07.008
networks: Surface and tribological properties. Polymer, [64] Dill K. A., Zimm B. H.: A rhelogical separator for very
51, 6219–6226 (2010). large DNA molecules. Nucleic Acids Research, 7, 735–
https://doi.org/10.1016/j.polymer.2010.10.036 749 (1979).
[54] Chen Z., Ward R., Tian Y., Eppler A. S., Shen Y. R., So- https://doi.org/10.1093/nar/7.3.735
morjai G. A.: Surface composition of biopolymer blends [65] MacDonald M. J., Muller S. J.: Experimental study of
biospan-SP/phenoxy and biospan-F/phenoxy observed shear-induced migration of polymers in dilute solutions.
with SFG, XPS, and contact angle goniometry. The Jour- Journal of Rheology, 40, 259–283 (1996).
nal of Physical Chemistry B, 103, 2935–2942 (1999). https://doi.org/10.1122/1.550740
https://doi.org/10.1021/jp984502z [66] Khan M. B., Briscoe B. J., Richardson M.: Field-in-
duced phase fractionation in multiphase polymer flow
systems: A review. Polymer-Plastics Technology and
Engineering, 33, 295–322 (1994).
https://doi.org/10.1080/03602559408013095

779