Accepted Manuscript Title: Effect of structure and viscosity of the components on some properties of starch-rich hybrid blends Author: Willian H. Ferreira Marwin M.I.B. Carmo Ana L´ucia N. Silva Cristina T. Andrade PII: DOI: Reference:
S0144-8617(14)01023-6 http://dx.doi.org/doi:10.1016/j.carbpol.2014.10.018 CARP 9371
To appear in: Received date: Revised date: Accepted date:
19-8-2014 11-10-2014 13-10-2014
Please cite this article as: Ferreira, W. H., Carmo, M. M. I. B., Silva, A. L. N., and Andrade, C. T.,Effect of structure and viscosity of the components on some properties of starch-rich hybrid blends, Carbohydrate Polymers (2014), http://dx.doi.org/10.1016/j.carbpol.2014.10.018 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3
Effect of structure and viscosity of the components on some properties of starch-rich
5
hybrid blends
ip t
4
cr
6
Willian H. Ferreira, Marwin M. I. B. Carmo, Ana Lúcia N. Silva, Cristina T. Andrade*
us
7 8
Instituto de Macromoléculas Professora Eloisa Mano, Universidade Federal do Rio de
10
Janeiro, Centro de Tecnologia, Bloco J, P.O. Box 68525, 21941-598, Rio de Janeiro, RJ,
11
Brazil
M
an
9
12
d
13
te
14
Ac ce p
15 16 17 18 19 20 21 22
*Corresponding author
23
Telephone number: + 55 21 2562 7208
24
Fax number: + 55 21 2270 1317
25
E-mail address: [email protected] 1
Page 1 of 32
26
Abstract Glycerol-plasticized cornstarch and poly(lactic acid) (PLA) were melt-blended alone and
28
at a constant 70:30 (m/m) composition, in the present of an organoclay. The effect of
29
increasing contents of the organoclay on extruded and compression-molded samples was
30
evaluated by X-ray diffraction (XRD), capillary rheometry, thermogravimetric analysis
31
(TGA), scanning electron microscopy (SEM), and tensile tests. XRD and shear viscosity
32
results obtained for the hybrid components (TPS/organoclay and PLA/organoclay) were
33
correlated with the hybrid blends properties. XRD and TGA results suggested that the
34
organoclay was similarly dispersed within both phases. SEM images revealed improved
35
adhesion
36
compatibilization as the organoclay content was increased. Some of the extruded materials
37
were also submitted to injection molding, and characterized by SEM and by tensile tests.
38
For the extruded and compression-molded samples, improved mechanical properties were
39
obtained for the samples with higher contents of the organoclay. For the injection-molded
40
samples, the mechanical properties seemed to be dependent on the organoclay dispersion.
phases.
Shear
viscosities
results
indicated
improved
te
d
M
the
Ac ce p
41 42
between
an
us
cr
ip t
27
Keywords: Thermoplastic starch; Poly(lactic acid); Hybrid blends; Extrusion, Injection molding
43 44
Chemical compounds studied in this article
45
Starch (PubChem CID: 24836924); Poly(lactic acid) (not found)
46 47 48
1. Introduction Poly(lactic acid) (PLA) and starch are considered promising biobased materials to
49 50
substitute traditional petroleum-based polymers. PLA is a biocompatible and bioresorbable 2
Page 2 of 32
polyester, with mechanical properties similar to poly(ethylene terephathalate) and
52
polystyrene. Until the late 1980s, the application of PLA was mainly focused on the
53
biomedical field, as a biomaterial for hard tissue replacement or in controlled drug delivery
54
systems. However, the high hydrophobic character and, consequently, slow hydrolytic
55
degradation rate, was appointed as a disadvantage to some applications (Luckachan &
56
Pillai, 2011).
cr
ip t
51
The interest in starch-based materials has grown because of their relatively low cost,
us
57
renewability and biodegradability in most environments. Starch consists of two
59
polysaccharides, amylose and amylopectin, both formed by D-glucose repeating units. The
60
semicrystalline structure of starch granules is destroyed by thermomecanical treatment in
61
the presence of suitable plasticizers, leading to thermoplastic starch (TPS). However, the
62
sensitivity of TPS to ambient humidity practically invalidates its commercial use. Because
63
of this drawback, the development of TPS blends and composites has been investigated.
64
Blending TPS with PLA would maintain the biodegradability of the resulting materials,
65
and might improve their properties.
Ac ce p
te
d
M
an
58
It was postulated that blend properties are determined by the structural characteristics of
66 67
each component and the phase morphology developed during melt processing. The final
68
morphology is controlled by blend composition, rheological properties of the components,
69
interfacial tension and viscosity ratio between the components, and processing conditions
70
(Sundararaj & Macosko, 1995; Wu, 1987). Much effort has been made to develop TPS/PLA blends. The low interfacial adhesion
71 72
between hydrophilic starch and hydrophobic PLA has been recognized as the major
73
problem to be overcome. To promote the compatibility of these materials, maleic
74
anhydride was added in one or two stages (Huneault & Li, 2007; Wang, Yu, & Ma, 2007;
75
Zhang & Sun, 2004). On the other hand, some authors followed three strategies; in situ 3
Page 3 of 32
formation of urethane linkages, in situ coupling of starch and PLA with peroxide, and
77
addition of previously formed PLA-grafted amylose as compatibilizing agent (Schwarch,
78
Six, & Avérous, 2008). Improvements in mechanical properties were obtained for the
79
blends with PLA in excess, substituting PLA for maleated PLA or by using PLA-grafted
80
amylose. Additionally, to prepare PLA/TPS blends, other authors used maleated PLA and
81
sodium montmorillonite (MMT) at 2% content, as a compatibilizing and a reinforcing
82
agent, respectively. Before processing, MMT was added either to a starch water-glycerol
83
suspension or to PLA. For the blends with 60% TPS, the location of MMT at the TPS
84
phase led to higher modulus and strength enhancement (Arroyo, Huneault, Favis, &
85
Bureau, 2010). In another report, small amounts of MMT (0.5 and 1% based on potato
86
starch dry mass) were also incorporated to a TPS/PLA blend at 60/40 composition, in a
87
multi-stage process, which consisted of potato starch gelatinization, ultrasonic dispersion
88
of MMT in water and melt mixing at 185ºC for 10 min. Reduction in water absorption and
89
improvements in mechanical and dynamic-mechanical properties were observed (Ayana,
90
Suin, & Khatua, 2014).
Ac ce p
te
d
M
an
us
cr
ip t
76
Similarly to the stabilization of foams and oil-in-water emulsions (Horozov, 2008;
91 92
Hunter, Pugh, Franks, & Jameson, 2008), nanoparticles may inhibit coalescence of
93
dispersed phase droplets in immiscible polymer blends, mainly when located at their
94
interface. However, the location of nanoparticles is dictated by thermodynamic
95
interactions, kinetic and viscosity effects (Feng, Chan, & Li, 2003; Fenouillot, Cassagnau,
96
& Majesté, 2009). Also, as postulated previously by these authors, the stabilization of
97
morphology is a requisite to maintain constant properties (Fenouillot at al., 2009). In this work, glycerol-plasticized cornstarch was melt-blended with PLA at a 70:30
98 99 100
constant composition without the addition of a compatibilizing agent, under one-step extrusion processing. Differently from the works cited before (Arroyo et al., 2010; Ayana 4
Page 4 of 32
et al., 2014), an organically-modified clay was also incorporated with increasing contents
102
(1.0 to 10.0 mass%) to the polymeric components. Both polymeric components were also
103
extruded and compression-molded alone and in the presence of the organoclay at the same
104
contents, and were characterized by X-ray diffraction and capillary rheometry. The
105
objectives of these experiments were to investigate the organoclay dispersion within TPS
106
and PLA matrices, determine changes in flow behavior resulting from the filler addition,
107
and mainly investigate the influence of these features on some properties of the resulting
108
hybrid blends. For this, TPS/PLA/organoclay blended materials were also characterized by
109
thermogravimetric analysis. The mechanical and morphological properties were
110
investigated for the extruded and compression-molded samples and for some of the
111
extruded materials , which were injection-molded .
112
Experimental
113
1.1 Materials
d
M
an
us
cr
ip t
101
Regular cornstarch (CS) was supplied by Ingredion Brasil - Ingredientes Industriais
te
114
Ltda. (São Paulo, SP, Brazil). According to the producer, this material is composed of 26–
116
30 mass% amylose and 74–70 mass% amylopectin, with less than 0.5 mass% gluten. The
117
gravimetric method was used to determine the humidity content (12 mass%). Analytical
118
grade glycerol was purchased from Vetec Química Fina (Rio de Janeiro, RJ, Brazil) and
119
was used as received. Poly(lactic acid) (PLA, NatureWorks LLC, Ingeo 2002D, with melt
120
flow index of 4-8 g/10min at 190°C / 2.16 kg) was supplied as pellets by Cargill Agrícola
121
S.A. (São Paulo, SP, Brazil). This PLA sample was previously characterized by gel
122
permeation chromatography in chloroform, and the results reported as Mn = 121.100 and
123
Mw = 197.000 (Souza, Dahmouche, Andrade, & Dias, 2013). PLA pellets were
124
cryogenically ground into powder to ensure better mixing with the other components.
125
Cloisite® 30B (C30B), a montmorillonite modified with dihydroxyethyl alkyl
Ac ce p
115
5
Page 5 of 32
126
methylammonium ions, was supplied by Southern Clay Products (Gonzales, TX, USA),
127
and was used as received.
128
1.2 Preparation of samples The mixture of CS and glycerol, added at 25 mass% based on starch dry mass, was
ip t
129
homogenized in a conventional mixer (Ika Works, Wilmington, NC, USA) for 30 min, and
131
maintained in tightly sealed polyethylene bags for 48 h at 4°C before processing to obtain
132
TPS. Glycerol-plasticized CS and PLA were mixed separately with C30B at 1.0, 2.5, 5.0,
133
7.5 and 10.0 mass% contents to obtain TPS/C30B and PLA/C30B materials. To obtain
134
TPS/PLA/C30B hybrid blends, glycerol-plasticized CS and PLA were homogenized at a
135
constant 70:30 (m/m) ratio and the organoclay was added at 1.0, 2.5, 5.0, 7.5 and 10.0
136
mass% contents, over the total mass of the polymers.
137
2.3. Extrusion and compression molding
M
an
us
cr
130
Extrusion processings were carried out in a Coperion ZSK 18 (Werner & Pfleiderer,
139
Stuttgart, Germany) co-rotating twin-screw extruder, with a L/D ratio of 40 and seven
140
heating zones. For glycerol-plasticized CS with various amounts of C30B, the seven
141
heating zones were maintained at 110, 110, 115, 115, 115, 110, 110°C, and the screw
142
speed was set at 120 rpm. For PLA and the blends (obtained by one-step extrusion
143
processings) with varying amounts of C30B, the seven heating zones were maintained at
144
160, 160, 165, 165, 165, 160, 160°C, and the screw speed was set at 120 rpm. The
145
PLA/C30B materials showed free-flowing behavior and were cooled in a water bath after
146
extrusion. All extruded materials were pelletized and compression-molded by heating at
147
different temperatures (TPS at 120ºC, PLA and the hybrid blends at 165ºC) under 68.9
148
MPa for 10 min, and cooling for 5 min in a cold press. The resulting sheets, with 12 cm x
149
10 cm x 2 cm dimensions, were used for the materials characterization.
150
2.4. X-ray diffraction (XRD)
Ac ce p
te
d
138
6
Page 6 of 32
The samples were analyzed with a Ultima IV diffractometer (Rigaku Corporation) in the
151
angular region 0.9 to 10° (2θ) at a 0.01º/s rate, in reflection mode. The CuKα1 radiation
153
with wavelength of 1.542 Å was generated at 40 kV and 20 mA. The d-values were
154
calculated by the Bragg's law (Equation 1). nλ = 2 d sen θ
155
Equation 1
ip t
152
where n is an integer, λ is the wavelength of incident wave (1.5418 Å), d is the spacing
157
between any two atomic planes in the crystal lattice, and θ is the angle of reflection.
158
2.5. Capillary viscosity measurements
us
cr
156
A Göttffert capillary rheometer model Rheograph 25 (Buchen, Germany), equipped
an
159
with a round-hole capillar of D = 1 mm (L/D = 30), was used to measure the flow
161
properties of TPS/C30B, PLA/C30B extruded materials and the hybrid blends. All tests
162
were performed over a shear rate range of 50 – 3000 s-1, at 165ºC with a prior melting
163
time of 5 min. End corrections were not applied; thus, the viscosity values are reported as
164
apparent viscosities. The setting of parameters and evaluation of raw data, as well as the
165
Rabinowitsch’s correction, were performed with the software LabRheo, provided with the
166
equipment. The consistency index and the flow behavior index were obtained according to
167
the power law model (Equations 2 and 3).
Ac ce p
te
d
M
160
168
169
τ=K
Equation 2
η=K
Equation 3
170
where τ is the shear stress, is the shear rate at the capillary wall, η is the viscosity, K is
171
the consistency index and n is the flow behavior index of the materials analyzed.
172
2.6. Thermogravimetric analysis
7
Page 7 of 32
Thermogravimetric analyses were carried out for the extruded materials under nitrogen
173
atmosphere on a TGA Q-500 equipment from TA Instruments (New Castle, USA).
175
Approximately, 10 mg of sample were heated from 25 to 700ºC, at a 10ºC/min rate.
176
2.7. Phase morphology
ip t
174
The phase morphology of the samples was examined with a Jeol electron microscope,
178
model JSM-6460LV (Akishima-shi, Japan), generally at the acceleration voltage of 20 kV.
179
Samples were cryogenically fractured before observation. The fractured surfaces were
180
vacuum-coated with gold before measurements. X-ray photoelectron spectroscopy (XPS)
181
was used to analyze the cryofractured surfaces of injection-molded samples in a
182
NORAN™ System 6 X-ray Microanalysis System, model C10015 (Thermo Fisher
183
Scientific Inc., Middleton, WI, USA), linked to the Jeol electron microscope.
184
2.8. Tensile tests of extruded and compression-molded samples
M
an
us
cr
177
Tensile tests were carried out at 21ºC in a Q-800 DMA equipment from TA Instruments
d
185
(New Castle, DE, USA), according to the ISO 527-3 method, with extruded and
187
compression-molded specimens with 13 ± 0.2 mm x 6 ± 0.2 mm x 0.8 ± 0.1 mm
188
dimensions, conditioned at 21ºC, and 50% relative humidity, for a period of at least 48 h.
189
The average value from a total of three measurements was taken. The samples were placed
190
between a fixed and a moveable clamp, and the experiments were performed in tensile
191
mode.
192
2.9. Injection molding
Ac ce p
te
186
The pelletized samples were subsequently dried in an oven at 100ºC for 2 h. Samples for
193 194
tensile tests were prepared in an injection-molding machine Arburg, model Allrounder
195
270S ( Lossburg, Germany) to obtain specimens according to ASTM D638, type I
196
(thickness 1.8 ± 0.1 mm). The five heating zones of the injection molding machine were
197
maintained at 170, 180, 190 and 200°C and 210ºC (from feed to die zone). The mold 8
Page 8 of 32
temperature was set at 30°C. The injection and the back pressures were 90 MPa and 57
199
MPa, respectively. The injection speed was 6 x 10-5 m3/s and the back time was 4 s. Before
200
performing mechanical tests, the specimens were conditioned at 21ºC and 50% relative
201
humidity for 48 h.
202
2.10. Tensile tests of extruded and injection-molded samples
ip t
198
Tensile tests were performed in a Instron Universal Testing Machine model 4204
204
(Norwood, MA, USA) equipped with a 1 kN load cell, at a speed of 1 mm/min. The
205
average value from a total of seven measurements was taken. All data were recorded and
206
processed by the software provided with the Instron equipment.
207
Results and discussion
an
us
cr
203
A diffractometer with a high resolution at low angles was used to investigate C30B
M
208
dispersion within TPS and PLA extruded materials. In Figure 1(a-c) (trace I), the 001
210
reflection for the neat C30B was observed at 2θ = 4.9º, and revealed a d001 value of 1.8 nm.
211
For TPS/C30B materials up to 5.0 mass% organoclay, no reflection was detected within
212
the 0.9 – 10º (2θ ) region (Figure 1a). This result indicated that the 001 reflection of C30B
213
was shifted to angles bellow 0.9° (2θ), with interlayer spacings higher than 9.8 nm. For the
214
TPS samples with C30B at 7.5 and 10 mass% contents, broad reflections were observed
215
around 4.5º and 4.3º (2θ), corresponding to d-values of approximately 2.0 nm. This result
216
may be attributed to the presence of a fraction of C30B aggregates, in which starch
217
molecules were partially intercalated. Although some authors reported on the
218
incompatibility of hydrophilic thermoplastic starch with C30B (Chiou et al., 2006), the
219
results of this work may be compared to those obtained by compounding potato starch with
220
C30B in an internal mixer (Park, Lee, Park, Cho, & Ha, 2003). In this latter work, no
221
reflection was found when C30B was incorporated at 2.5 mass% content, and a slight shift
222
in the organoclay 001 reflection, from around 4.71º (2θ) to around 4.30º (2θ), was
Ac ce p
te
d
209
9
Page 9 of 32
observed for the hybrids with 5 and 10 mass% C30B. The presence of two hydroxyethyl
224
moieties in the intercalant of C30B explains its interaction with starch molecules. On the
225
other hand, for all PLA/C30B materials (Figure 1b), low-intensity reflections around 2.6º
226
(2θ) were observed. The intensity of these reflections increased slightly with the increase
227
in organoclay content, and indicated that the organoclay had interlayer spacings of
228
approximately 3.4 nm, intercalated with PLA molecules. Similar results were obtained by
229
other authors for PLA and C30B (Araújo, Botelho, Oliveira, & Machado, 2014; Duan,
230
Thomas, & Huang, 2013; Fukushima, Tabuani, & Camino, 2012; Lai, Wu, Lin, & Don,
231
2014; Molinaro et al., 2013; Pluta, 2006; Zaidi et al., 2010), and attributed to hydrogen
232
bonding interaction between the hydroxyethyl groups of the oraganoclay intercalant and
233
carbonyl ester groups of PLA macromolecules. Anyway, the difference between PLA and
234
TPS intercalation might be ascribed to the bulky ring structure of TPS, because in both
235
cases hydrogen bonding interactions are favored.
d
M
an
us
cr
ip t
223
It would be worth observing the diffratograms obtained for the TPS/PLA/C30B
te
236
materials. In Figure 1c (traces II – IV), following the results for TPS/C30B, no reflection
238
was observed in the diffratograms for TPS/PLA hybrids with 1.0, 2.5 and 5.0 mass%
239
C30B, which indicated that the 001 reflection of C30B was shifted to angles bellow 0.9°
240
(2θ), with interlayer spacings higher than 9.8 nm. For the hybrids with 7.5 and 10.0 mass%
241
C30B, shallow reflections around 3º and 4.5 - 5º (2θ) were envisaged. In this case, the
242
reflection around 3º (2θ) was attributed to C30B layers intercalated with starch molecules,
243
and the other broader and less intense reflection at 4.5 – 5º (2θ) was attributed to a very
244
small amount of C30B aggregates. Anyway, these diffractograms corroborated the good
245
affinity of C30B towards both TPS and PLA phases, and suggested an even better
246
dispersion of the organoclay in the extruded blends in relation to TPS or PLA alone.
Ac ce p
237
10
Page 10 of 32
The melt flow characteristics of all extruded materials considered in this work were
248
investigated by capillary rheometry at 165ºC, and the resulting curves are shown in Figure
249
2 (a-c), as apparent viscosity versus shear rate. Shear thinning behavior was observed for
250
all samples. Comparing TPS and PLA alone, the virgin PLA presented higher viscosity
251
values, but a less pronounced shear thinning behavior compared to TPS, as indicated by the
252
higher flow index (n) value, shown in Table 1. However, for the extruded PLA sample, a
253
severe reduction of viscosity was observed. In general, processing of PLA is preceded by
254
humidity elimination, but even when this procedure was followed, degradation was
255
observed (Souza et al., 2013). Viscosity reduction of PLA was attributed to both hydrolysis
256
reaction and thermal degradation (Speranza, De Meo, & Pantani, 2014). Since the
257
objective of the extrusion processing was to produce TPS/PLA hybrid blends, in which all
258
components are hygroscopic, drying seemed an unrealistic procedure.
M
an
us
cr
ip t
247
For TPS/C30B materials (Figure 2a), the increase in organoclay content, from 1.0 to 10
d
259
mass% content, led to almost a half-decade increase of apparent viscosity, and indicated
261
that both exfoliated (at a low C30B content), and particularly intercalated organoclay
262
contributed positively to viscosity. For these materials, the flow index n (Table 1) showed
263
typical values, within the range (0.27 – 0.56) previously reported for cornstarch submitted
264
to thermomechanical treatments at different moisture contents (Vergnes & Villemaire,
265
1987). On the other hand, a substantial reduction of viscosity and n values was observed
266
for PLA/C30B materials with increasing organoclay content (Figure 2b, Table 1).
267
Recently, the thermal decomposition process of a similar system composed of PLA and
268
C30B at 0.5 and 2.5 mass% content was reported, and the nanocomposites were found
269
more stable than PLA alone (Carrasco, Pérez-Maqueda, Santana, & Maspoch, 2014). In the
270
last case, a dehumidifier was used prior to extrusion. When comparing the two hybrids, the
271
authors observed that the incorporation of 2.5 mass% C30B to PLA led to a less thermally
Ac ce p
te
260
11
Page 11 of 32
stable sample, attributed to the presence of C30B aggregates. In this study, the
273
diffractograms of Figure 1b revealed that the extrusion conditions led to intercalated C30B
274
hybrids. Thus, the reduction of viscosity, mainly at 7.5 and 10.0 mass% C30B contents,
275
and at high shear rates, might be indicating hydrolytic degradation to such an extent that
276
topological constraints were extensively minimized.
ip t
272
Viscosity curves were also obtained for the 70:30 TPS/PLA blends, with or without the
cr
277
addition of C30B (Figure 2c, Table 1). As expected, the melt viscosity for the neat
279
TPS/PLA blend resulted in intermediary values between those determined for TPS and
280
PLA, and the incorporation of C30B contributed to increase the blends viscosity. The flow
281
index values seemed to be governed by the properties of TPS, the component added at a
282
larger composition. However, no clear trend was followed by the consistency index K. It
283
might be inferred that the more pronounced Newtonian character of the TPS/PLA hybrid
284
with C30B at 10.0 mass% content (Table 1) would worsen conditions for further
285
processing, such as injection molding.
te
d
M
an
us
278
The thermal stability of TPS, PLA, and TPS/PLA blends with and without the addition
287
of C30B was evaluated by thermogravimetric analysis (TGA) under nitrogen atmosphere.
288
The TGA curves are shown in Figure 3 (a,b). Up to 100ºC, mass loss is mainly attributed
289
to humidity loss. From 100ºC to decomposition temperature, the mass loss observed for
290
TPS and TPS/PLA/C30B at 1.0 mass% might be attributed to glycerol evaporation. With
291
increasing organoclay content, a reduction of plasticizer loss was observed. The neat
292
TPS/PLA blend showed the lowest onset degradation temperature (To = 284ºC), whereas
293
TPS (To = 297ºC) and PLA (To = 332ºC) had higher thermal stabilities. For the hybrid
294
blends, To decreased moderately, from 303ºC to 288ºC, as the C30B content was increased.
295
It is worth observing in Figure 3 (c,d) (DTG curves) that the maximum degradation rate
296
was reached for PLA at Tmax = 361ºC, and for TPS at Tmax = 327ºC. While a single peak at
Ac ce p
286
12
Page 12 of 32
Tmax = 301ºC was observed for the neat TPS/PLA blend, two peaks were observed for the
298
hybrid samples. The first peak, attributed to the maximum degradation rate of the TPS
299
phase, varied from Tmax = 302ºC to Tmax = 313ºC with increasing C30B content. Lower
300
Tmax values for TPS/calcium montmorillonite hybrids in relation to TPS were previously
301
observed, and were attributed to the catalytic effect of acidic sites of montmorillonite on
302
the thermal degradation of the starch matrix (Magalhães & Andrade, 2010). Similar result
303
was also observed for other hybrid systems (Ramos Filho, Mélo, Rabello, & Silva, 2005).
304
The second peak, attributed to the maximum degradation rate of the PLA phase, varied
305
from Tmax = 357ºC to Tmax = 369ºC with increasing C30B content. This result is in
306
accordance with some authors, to whom the preparation of nanocomposites by adding
307
organo-modified montmorillonites should be considered as a promising method to reduce
308
PLA thermal degradation (Carrasco et al, 2014). In relation to PLA, other authors observed
309
a slight increase in Tmax for the PLA/C30B nanocomposites at 1.0, 3.0 and 5.0 phr
310
organoclay content (Lai et al., 2014).
te
d
M
an
us
cr
ip t
297
In this case, the TGA data also corroborated viscosity results for the TPS/PLA/C30B
312
hybrids (Figure 2), which suggested the protecting action of TPS against PLA degradation.
313
More important, DTG curves gave additional evidence on the location of clay mineral
314
particles in both TPS and PLA phases, as suggested by XRD results (Figure 1). Also, as
315
observed by some authors (Lai et al., 2014), the nucleating effect of the organoclay
316
particles should not be neglected. XRD data in the 2θ region of 2º to 35º revealed PLA
317
characteristic reflections with increasing intensities, as the content of C30B was increased
318
(Figure 1, Supplementary data).
Ac ce p
311
As in most polymer blends, TPS and PLA are immiscible. Melt blending was performed
319 320
without the addition of an organic compatibilizer and, as expected, the image of Figure 4a,
321
obtained for the neat TPS/PLA blend, shows a coarse morphology. Immiscibility was 13
Page 13 of 32
evidenced by the size, and the lack of adhesion between the two phases. For the extruded
323
and compression-molded TPS/PLA/C30B materials in Figure 4 (b-f), the micrographs
324
revealed that, as the organoclay content was increased from 2.5 to 10 mass%, at 10
325
mass% the PLA dispersed phase became undistinguished from the TPS matrix. This
326
result was related to the role of the organoclay in the improvement of the system
327
morphology. . Considering that the organoclay interacted similarly with both components,
328
and considering the viscosity results obtained, a rough estimation of viscosity ratios might
329
be envisaged for the 70:30 TPS/PLA blends with added C30B. With increasing C30B
330
contents, the viscosity ratio p and the capillary number might be expected to decrease, and
331
p might approach the value p = 1. Contrarily to the morphology improvement, previously
332
reported for C30B nanocomposites of TPS and poly(3-hydroxybutyrate-co-3-
333
hydroxyvalerate) (PHBV) (Magalhães, Dahmouche, Lopes, & Andrade, 2013), no clear
334
visualization of size reduction of the PLA dispersed phase was observed, as the content of
335
C30B was increased from 2.5 to 10 mass%. Thus, no indication on the location of the
336
organoclay nanoparticles at the interface between the components could be suggested.
337
Also, it was claimed that clay mineral particles would be incorporated preferentially into
338
the polymer with the lowest melting temperature (Fenouillot et al., 2009). In this case, TPS
339
has a lower melting temperature than PLA. However, as shown in Figure 2 for TPS/C30B
340
and PLA/C30B hybrids, with increasing contents of the organoclay, while the viscosity of
341
TPS was increased, the viscosity of PLA was decreased, and the melting pattern might be
342
changed even during the extrusion step. As the viscosity of the polymeric components
343
seems to interfere on the location of the clay minerals particles (Naderi, Lafleur, & Dubois,
344
2008), it might be presumed that C30B tended to be almost equally incorporated in both
345
TPS and PLA phases during processing, and eventually at the interface, as its content was
346
increased. However, it seems as if the amount of organoclay was not enough to saturate
Ac ce p
te
d
M
an
us
cr
ip t
322
14
Page 14 of 32
347
the interface between the TPS matrix and PLA droplets, probably because exfoliated
348
organoclay layers or organoclay aggregates were preferentially located within both
349
polymers. The mechanical properties of the TPS/PLA hybrid blends were investigated by two
351
methods, using compression-molded (Figure 5) and injection-molded specimens (Figure
352
6). These results were analyzed independently, since they were acquired by different
353
methods. In Figure 5a, for compression-molded samples, the rigid character of PLA, to
354
which the elastic modulus was determined as E = 3146.3 ± 311.8 MPa, and the soft nature
355
of TPS were once more revealed. In Figure 5b, even for the neat TPS/PLA blend, the
356
elastic modulus (190.2 ± 18.1 MPa) increased in relation to TPS alone (E = 82.3 ± 7.9
357
MPa). For the compression-molded hybrids, the moduli varied from E = 167.5 ± 14.8 MPa
358
to 501.4 ± 52.2 MPa, by incorporating 1.0 to 7.5 mass% C30B. This result was expected
359
because of the rigid nature of the filler. However, a substantial decrease in tensile stress at
360
break (σmax) was observed for the neat TPS/PLA blend (σmax = 1.5 ± 0.1 MPa) and the
361
C30B hybrids, in relation to TPS and PLA. For TPS, this value was σmax = 4.4 ± 0.4 MPa
362
and, for PLA, σmax = 5.3 ± 0.5 MPa. For the TPS/PLA hybrids, σmax varied from 1.7 ± 0.2
363
MPa for C30B incorporated at a 1.0 mass% content to σmax = 3.4 ± 0.4 MPa for the hybrid
364
with C30B at a 7.5 mass% content. As for the elongation at break, εmax, their values varied
365
for the same hybrids, respectively, from 0.8 ± 0.1% to 1.3 ± 0.01%, in relation to εmax =
366
15.2 ± 1.1% for TPS.
Ac ce p
te
d
M
an
us
cr
ip t
350
For the hybrid prepared with C30B at 10.0 mass% content, the E and σmax values
367 368
decreased to E = 383.5 ± 41.2 MPa and σmax = 2.9 ± 0.3 MPa, respectively. This result was
369
not expected, as long as increasing values were obtained with increasing contents of C30B.
370
However, the plasticizing effect of the C30B intercalant, also incorporated at a higher 15
Page 15 of 32
371
amount, might have contributed to the less rigid character of this hybrid, for which εmax
372
reached 1.9 ± 0.01%. It is worth noting that the relatively better results of tensile tests (in relation to the neat
373
TPS/PLA blend) for the hybrids, to which C30B was incorporated at 5.0 to 10.0 mass%
375
contents, followed the previously rough estimation for viscosity ratios. Because of this
376
result, these extruded samples as well as the neat 70:30 TPS/PLA blend were chosen to be
377
submitted to injection molding, to investigate possible changes of morphology and
378
mechanical properties.
us
cr
ip t
374
The mechanical properties of the injection-molded samples were evaluated (Figure 6).
380
The highest stress at break (σmax = 18.9 MPa) was obtained for the neat TPS/PLA blend.
381
Differently from the previous result obtained for the compression-molded sample, the
382
injection-molded hybrid with 10.0 mass% C30B did not show the highest ductibility. The
383
highest elastic modulus (E = 11.4 ± 3.8 MPa) and lowest elongation at break (εmax = 0.55 ±
384
0.16%) were determined for the hybrid with 7.5 mass% C30B.
te
d
M
an
379
The best ductibility was achieved for the TPS/PLA/C30B hybrid with 5.0 mass%
Ac ce p
385 386
organoclay (E = 1091.5 ± 181.3 MPa, σmax = 11.2 ± 2.2 MPa, εmax = 2.33 ± 1.4%). Bearing
387
in mind XRD results, the incorporation of 5.0 mass% C30B led to a higher degree of
388
exfoliation, which certainly contributed to the better fluidity during injection molding and
389
the highest ductibility. Moreover, the more pronounced Newtonian character of the
390
TPS/PLA hybrid with C30B at 10.0 mass%, as shown before (Table 1), indeed made
391
injection molding more difficult. To verify the effect of injection molding on morphology, cryofractured surfaces of the
392 393
injection-molded samples were analyzed by SEM (Figure 7). Although more sensitive to
394
the electron beam, the neat TPS/PLA blend in Figure 7a showed a similar image to that
395
taken for the compression-molded sample (Figure 4a). The SEM images for the hybrids 16
Page 16 of 32
with 5.0 and 7.5 mass% C30B in Figure 7 (b,c) were characterized by the appearance of
397
microvoids much larger in diameter than those observed for the corresponding
398
compression-molded samples. The appearance of these microvoids might be related to the
399
loss of some droplets of PLA dispersed phase during cryofracturing. If this supposition
400
were true, it indicated a very small amount of C30B at the interface. No particular feature
401
seems to distinguish the image observed for the injection-molded hybrid with C30B at 10
402
mass% content from that obtained for the compression-molded sample with the same
403
content of the filler. This result might be probably related to a higher amount of the
404
organoclay at the interface.
an
us
cr
ip t
396
Using XPS to analyze the cryofractured surfaces of injection-molded samples, the
406
results revealed widespread distribution of Mg, Al and Si (Supplementary data, Figures 2-
407
4), according to the previous supposition on incorporation of C30B within both
408
components phases.
409
Conclusions
te
d
M
405
The favorable interaction between the organoclay intercalant and the molecules of both
Ac ce p
410 411
components, starch and PLA, suggested by XRD and TGA results, seemed to dictate the
412
properties of extruded TPS/PLA/C30B hybrids. XRD data suggested no specific
413
preference for one of the blend components. TGA indicated the stabilizing effect of C30B
414
on the PLA phase, as the content of the organoclay increased. Shear viscosities,
415
determined for each component in the presence of the same organoclay contents, roughly
416
indicated the tendency for achieving a viscosity ratio around p = 1, as higher contents of
417
organoclay were incorporated. For the extruded and compression-molded samples, the
418
results of tensile tests followed this estimation. The SEM images revealed that the
419
morphology, even after injection molding, remained the same for the TPS/PLA hybrid
420
blend, prepared with C30B at the highest content (10.0 mass%). The high degree of 17
Page 17 of 32
organoclay dispersion, associated with the highest pseudoplasticity of the 5.0 mass%
422
hybrid blend, and the structuring characteristic of the process, certainly contributed to the
423
highest ductibility of this TPS/PLA hybrid material after processing by injection molding.
424
Acknowledgments
ip t
421
The authors acknowledge Conselho Nacional de Desenvolvimento Científico e
426
Tecnológico (CNPq), Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro
427
(FAPERJ), and Coordenação para o Aperfeiçoamento de Pessoal de Nível Superior
428
(CAPES) for financial support.
429
References
430
Araújo, A., Botelho, G., Oliveira, M., & Machado, A.V. (2014). Influence of clay organic
431
modifier on the thermal-stability of PLA based nanocomposites. Applied Clay Science,
432
88-89, 144-150.
us an
M
Arroyo, O. H., Huneault, M. A., Favis, B. D., Bureau, M. N. (2010). Processing and
d
433
cr
425
properties of PLA/thermoplastic starch/montmorillonite nanocomposites. Polymer
435
Composites, 31, 114–127.
Ac ce p
436
te
434
Ayana, B., Suin, S., & Khatua, B. B. (2014). Highly exfoliated eco-friendly thermoplastic starch (TPS)/poly (lactic acid)(PLA)/clay nanocomposites using unmodified
437
nanoclay. Carbohydrate Polymers, 110, 430–439.
438 439
Carrasco, F., Pérez-Maqueda, L. A., Santana, O.O., & Maspoch, M. Ll. (2014). Enhanced general analytical equation for the kinetics of the termal degradation of poly(lactic
440 441
acid)/montmorillonite nanocomposites driven by random scission. Polymer
442
Degradation and Stability, 101, 52-59.
443
Chiou, B.-S., Yee, E., Wood, D., Shey, J., Glenn, G., & Orts, W. (2006). Effects of
444
processing conditions on nanoclay dispersion in starch-clay nanocomposites. Cereal
445
Chemistry, 83, 300-305. 18
Page 18 of 32
446
Duan, Z., Thomas, N. L., & Huang, W. (2013). Water vapour permeability of poly(lactic acid) nanocomposites. Journal of Membrane Science, 445, 112-118.
447 448
Feng, J., Chan, C. M., & Li, J. X. (2003). A method to control the dispersion of carbon black in an immiscible polymer blend. Polymer Engineering and Science, 43, 1058–
450
1063.
Fenouillot, F., Cassagnau, P., & Majesté, J.-C. (2009). Uneven distribution of
cr
451
ip t
449
nanoparticles in immiscible fluids: morphology development in polymer blends.
453
Polymer, 50, 1333–1350.
Fukushima, K., Tabuani, D., & Camino, G. (2012). Poly(lactic acid)/clay nanocomposites:
an
454
us
452
effect of nature and content of clay on morphology, thermal and thermo-mechanical
456
properties. Materials Science and Engineering C 32, 1790-1795.
457
M
455
Horozov, T. S. (2008). Foams and foam films stabilised by solid particles. Current Opinion in Colloid & Interface Science, 13, 134–140. Huneault, M. A., Li, H. (2007). Morphology and properties of compatibilized
te
459
d
458
polylactide/thermoplastic starch blends. Polymer, 48, 270–280.
Ac ce p
460 461
Hunter, T. N., Pugh, R. J., Franks, G. V., & Jameson, G. J. (2008). The role of particles in
462
stabilising foams and emulsions. Advances in Colloid and Interface Science, 137, 57–
81.
463 464
Lai, S.-M., Wu, S.-H., Lin, G.-G, & Don, T.-M. (2014). Unusual mechanical properties of melt-blended poly(lactic acid) (PLA)/clay nanocomposites. European Polymer
465
Journal, 52, 193-206.
466 467
Luckachan, G.E., & Pillai, C.K.S. (2011). Biodegradable polymers – a review on recent
468
trends and emerging perspectives. Journal of Polymers and the Environment, 19, 637-
469
676.
19
Page 19 of 32
470
Magalhães, N. F., & Andrade, C. T. (2010). Calcium bentonite as reinforcing nanofiller for thermoplastic starch. Journal of the Brazilian Chemical Society, 21, 202-208.
471 472
Magalhães, N. F., Dahmouche, K., Lopes, G. K., & Andrade, C. T. (2013). Using an organically-modified montmorillonite to compatibilize a biodegradable blend. Applied
474
Clay Science, 72, 1–8.
ip t
473
Molinaro, S., Romero, M. C., Boaro, M., Sensidoni, A., Lagazio, C., Morris, M., & Kerry,
476
J. (2013). Effect of nanoclay-type and PLA optical purity on the characteristics of
477
PLA-based nanocomposite films. Journal of Food Engineering, 117, 113-123.
478
Naderi, G., Lafleur, P. G., & Dubois, C. (2008). The influence of matrix viscosity and
an
us
cr
475
composition on the morphology, rheology, and mechanical properties of thermoplastic
480
elastomer nanocomposites based on EPDM/PP. Polymer Composites, 29, 1301-1309. Park, H.-M., Lee, W., Park, C.-Y., Cho, W.-J., & Ha, C.-S. (2003). Environmentally
d
481
M
479
friendly hybrids. Part I Mechanical, thermal, and barrier properties of thermoplastic
483
starch/clay nanocomposites. Journal of Materials Science, 38, 909-915.
te
482
Pluta, M. (2006). Melt compounding of polylactide/organoclay: structure and properties of
485
nanocmposites. Journal of Polymer Science Part B: Polymer Physics, 44, 3392-3405.
Ac ce p
484
486
Ramos Filho, F. G., Mélo, T. J. A., Rabello, M. S., & Silva, S. M. L. (2005). Thermal
487
stability of nanocomposites based on polypropylene and bentonite. Polymer Degradation and Stability, 89, 383-392.
488 489
Schwach, E., Six, J.-L., & Avérous, L. (2008). Biodegradable blends based on starch and
490
poly(lactic acid): comparison of different strategies and estimate of compatibilization.
491
Journal of Polymers and the Environment, 16, 286–297.
20
Page 20 of 32
492
Souza, D. H. S., Dahmouche, K., Andrade, C. T., & Dias, M. L. (2013). Synthetic
493
organofluoromica/poly(lactic acid) nanocomposites: structure, rheological and thermal
494
properties. Applied Clay Science, 80-81, 259-266. Speranza, V., De Meo, A., & Pantani, R. (2014).Thermal and hydrolytic kinetics of PLA in the molten state. Polymer Degradation and Stability, 100, 37-41.
497
cr
496
ip t
495
Sundararaj, U., & C. W. Macosko, C. W. (1995). Drop breakup and coalescence in polymer blends: the effects of concentration and compatibilization. Macromolecules,
499
28, 2647-2657.
Vergnes, B., & Villemaire, J. P. (1987). Rheological behaviour of low moisture molten
an
500
maize starch. Rheological Acta, 26, 570-576.
501
Wang, N., Yu, J., & Ma, X. (2007). Preparation and characterization of thermoplastic
M
502
us
498
starch/PLA blends by one‐step reactive extrusion. Polymer International, 56, 1440–
Ac ce p
te
d
503
1447.
504 505
Wu, S. (1987). Formation of dispersed phase in incompatible polymer: interfacial and rheological effects. Polymer Engineering and Science, 27, 335-343.
506 507
Zaidi, L., Bruzaud, S., Bourmaud, A., Médéric, P., Kaci, M., & Grohens, Y. (2010). Journal of Applied Polymer Science, 116, 1357-1365.
508 509
Zhang, J. - F., & Sun, X. (2004). Mechanical properties of poly(lactic acid)/starch composites compatibilized by maleic anhydride. Biomacromoleules, 5, 1446-1451.
510 511 512
21
Page 21 of 32
513 514 515
ip t
516 517
cr
518
us
519 520
an
521 522
M
523 524
d
525
te
526
Ac ce p
527 528 529
Table 1: Power law constants for glycerol-plasticized thermoplastic starch, poly(acid acid)
530
and their blends, with or without the incorporation of the C30B.
531
K (Pa.s)
n
R2
TPS
10 399
0.27
0.998
TPS/ C30B (1.0 mass%)
9 727
0.31
0.999
TPS/ C30B (2.5 mass%)
9 332
0.36
0.999
TPS/ C30B (5.0 mass%)
13 489
0.31
0.999
Blend composition
22
Page 22 of 32
21 281
0.28
0.998
TPS/C30B (10.0 mass%)
25 577
0.31
0.999
PLA (as received)
44 668
0.40
0.995
964
0.32
0.996
PLA/C30B (1.0 mass%)
100 025
0.17
PLA/C30B (2.5 mass%)
51 286
0.25
PLA/C30B (5.0 mass%)
41 686
0.15
PLA/C30B (7.5 mass%)
125 892
PLA/C30B (10.0 mass%)
25 150
TPS/PLA
0.998
us
cr
0.995 0.996
0.999
-0.04
0.998
13 423
0.37
0.998
TPS/PLA/C30B (1.0 mass%)
25 118
0.29
0.997
TPS/PLA/C30B (2.5 mass%)
40 738
0.34
0.994
81 234
0.28
0.999
51 121
0.39
0.996
21 579
0.60
0.998
an
-0.44
M
PLA (after extrusion)
ip t
TPS/C30B (7.5 mass%)
te
TPS/PLA/C30B (7.5 mass%)
d
TPS/PLA/C30B (5.0 mass%)
TPS/PLA/C30B (10.0 mass%)
Ac ce p
532 533 534
Caption to the figures
535
Figure 1: X-ray diffractograms in the 2θ region of 0.6º to 10º for the neat C30B (traces I);
536
(a) TPS, (b) PLA, (c) 70:30 TPS/PLA blend with the incorporation of C30B at 1.0 (traces II), 2.5 (traces III), 5.0 (traces IV), 7.5 (traces V), 10.0 (traces VI) mass%.
537 538
Figure 2: Variation of apparent viscosity as a function of shear rate at 165ºC for
539
(a) TPS alone (■) and hybrids, (b) PLA alone (■) and hybrids, and (c) neat 70:30
540
TPS/PLA blend (■) and hybrids, with the incorporation of C30B at 1.0 (U), 2.5 ( ),
541
5.0 ({), 7.5 (V), and 10.0 () mass%.
23
Page 23 of 32
Figure 3: TGA curves for (a) TPS (¯) and PLA ( ), and (b) neat 70:30 TPS/PLA blend
543
(■) and hybrids, (c) DTG curves for TPS (¯) and PLA ( ), and (d) neat 70:30
544
TPS/PLA blend (■) and hybrids with the incorporation of C30B at 1.0% ( ), 2.5%
545
(U), 5.0 ({) 7.5 % (V), and 10.0 () mass%.
ip t
542
Figure 4: SEM images for the extruded samples; neat 70:30 TPS/PLA blend (a), and for
547
the hybrids with the incorporation of C30B at 1.0 (b), 2.5 (c), 5.0 (d), 7.5 (e) and 10.0
548
(f) mass%.
us
cr
546
Figure 5: Stress versus strain curves for the extruded samples; (a) TPS ( ) and PLA (¯);
550
(b) neat 70:30 TPS/PLA blend (■) and hybrids with the incorporation of C30B at 1.0%
551
( ), 2.5% (U), 5.0 ({) 7.5 % (V), and 10.0 () mass%.
an
549
Figure 6: Stress versus strain curves for the injection-molded samples; neat 70:30
553
TPS/PLA blend (■) and for the hybrids the incorporation of C30B at 5.0 ({) 7.5 %
554
(V), and 10.0 () mass%.
d
Figure 7: SEM images for the injection-molded samples; neat 70:30 TPS/PLA blend (a),
te
555
M
552
and for the hybrids the incorporation of C30B at 5.0 (b) 7.5 % (c), and 10.0 (d) mass%.
Ac ce p
556 557 558 559
24
Page 24 of 32
ip t cr M
an
us
560
Ac ce p
te
d
561
562 563 564
565 566 567 568
Figure 1
569
25
Page 25 of 32
cr
ip t
570
M
an
us
571
Ac ce p
te
d
572
573 574 575 576 577
Figure 2
26
Page 26 of 32
ip t M
an
us
cr
578 579
Ac ce p
te
d
580 581
582 583
584 585
Figure 3
27
Page 27 of 32
ip t cr M
an
us
586
Ac ce p
te
d
587
588
589 28
Page 28 of 32
590 591 592
ip t
Figure 4
an
us
cr
593
Ac ce p
te
d
M
594
595 596 597 598 599
600 601 602 603
29
Page 29 of 32
604 605
Figure 5
606
ip t
607
M
an
us
cr
608
609 610
te
d
611 612
Ac ce p
613 614 615 616 617 618 619 620 621 622 623
30
Page 30 of 32
Figure 6
cr
ip t
624
d
M
an
us
625
Ac ce p
te
626
627 628
629 fig. 7
31
Page 31 of 32
638
Highlights
639 640
• Melt-blending of cornstarch hybrid blends were investigated.
ip t
641
643
cr
642
• Hybrids of both components were prepared at the same contents of the organoclay.
645
us
644
• Results indicated the incorporation of the organoclay in both phases.
647
an
646
• .
• The organoclay dispersion and pseudoplasticity affected hybrid materials properties.
d
649
M
648
te
650 651
Ac ce p
652 653 654 655 656 657 658 659
660
33
Page 32 of 32
S0144-8617(14)01023-6 http://dx.doi.org/doi:10.1016/j.carbpol.2014.10.018 CARP 9371
To appear in: Received date: Revised date: Accepted date:
19-8-2014 11-10-2014 13-10-2014
Please cite this article as: Ferreira, W. H., Carmo, M. M. I. B., Silva, A. L. N., and Andrade, C. T.,Effect of structure and viscosity of the components on some properties of starch-rich hybrid blends, Carbohydrate Polymers (2014), http://dx.doi.org/10.1016/j.carbpol.2014.10.018 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1 2 3
Effect of structure and viscosity of the components on some properties of starch-rich
5
hybrid blends
ip t
4
cr
6
Willian H. Ferreira, Marwin M. I. B. Carmo, Ana Lúcia N. Silva, Cristina T. Andrade*
us
7 8
Instituto de Macromoléculas Professora Eloisa Mano, Universidade Federal do Rio de
10
Janeiro, Centro de Tecnologia, Bloco J, P.O. Box 68525, 21941-598, Rio de Janeiro, RJ,
11
Brazil
M
an
9
12
d
13
te
14
Ac ce p
15 16 17 18 19 20 21 22
*Corresponding author
23
Telephone number: + 55 21 2562 7208
24
Fax number: + 55 21 2270 1317
25
E-mail address: [email protected] 1
Page 1 of 32
26
Abstract Glycerol-plasticized cornstarch and poly(lactic acid) (PLA) were melt-blended alone and
28
at a constant 70:30 (m/m) composition, in the present of an organoclay. The effect of
29
increasing contents of the organoclay on extruded and compression-molded samples was
30
evaluated by X-ray diffraction (XRD), capillary rheometry, thermogravimetric analysis
31
(TGA), scanning electron microscopy (SEM), and tensile tests. XRD and shear viscosity
32
results obtained for the hybrid components (TPS/organoclay and PLA/organoclay) were
33
correlated with the hybrid blends properties. XRD and TGA results suggested that the
34
organoclay was similarly dispersed within both phases. SEM images revealed improved
35
adhesion
36
compatibilization as the organoclay content was increased. Some of the extruded materials
37
were also submitted to injection molding, and characterized by SEM and by tensile tests.
38
For the extruded and compression-molded samples, improved mechanical properties were
39
obtained for the samples with higher contents of the organoclay. For the injection-molded
40
samples, the mechanical properties seemed to be dependent on the organoclay dispersion.
phases.
Shear
viscosities
results
indicated
improved
te
d
M
the
Ac ce p
41 42
between
an
us
cr
ip t
27
Keywords: Thermoplastic starch; Poly(lactic acid); Hybrid blends; Extrusion, Injection molding
43 44
Chemical compounds studied in this article
45
Starch (PubChem CID: 24836924); Poly(lactic acid) (not found)
46 47 48
1. Introduction Poly(lactic acid) (PLA) and starch are considered promising biobased materials to
49 50
substitute traditional petroleum-based polymers. PLA is a biocompatible and bioresorbable 2
Page 2 of 32
polyester, with mechanical properties similar to poly(ethylene terephathalate) and
52
polystyrene. Until the late 1980s, the application of PLA was mainly focused on the
53
biomedical field, as a biomaterial for hard tissue replacement or in controlled drug delivery
54
systems. However, the high hydrophobic character and, consequently, slow hydrolytic
55
degradation rate, was appointed as a disadvantage to some applications (Luckachan &
56
Pillai, 2011).
cr
ip t
51
The interest in starch-based materials has grown because of their relatively low cost,
us
57
renewability and biodegradability in most environments. Starch consists of two
59
polysaccharides, amylose and amylopectin, both formed by D-glucose repeating units. The
60
semicrystalline structure of starch granules is destroyed by thermomecanical treatment in
61
the presence of suitable plasticizers, leading to thermoplastic starch (TPS). However, the
62
sensitivity of TPS to ambient humidity practically invalidates its commercial use. Because
63
of this drawback, the development of TPS blends and composites has been investigated.
64
Blending TPS with PLA would maintain the biodegradability of the resulting materials,
65
and might improve their properties.
Ac ce p
te
d
M
an
58
It was postulated that blend properties are determined by the structural characteristics of
66 67
each component and the phase morphology developed during melt processing. The final
68
morphology is controlled by blend composition, rheological properties of the components,
69
interfacial tension and viscosity ratio between the components, and processing conditions
70
(Sundararaj & Macosko, 1995; Wu, 1987). Much effort has been made to develop TPS/PLA blends. The low interfacial adhesion
71 72
between hydrophilic starch and hydrophobic PLA has been recognized as the major
73
problem to be overcome. To promote the compatibility of these materials, maleic
74
anhydride was added in one or two stages (Huneault & Li, 2007; Wang, Yu, & Ma, 2007;
75
Zhang & Sun, 2004). On the other hand, some authors followed three strategies; in situ 3
Page 3 of 32
formation of urethane linkages, in situ coupling of starch and PLA with peroxide, and
77
addition of previously formed PLA-grafted amylose as compatibilizing agent (Schwarch,
78
Six, & Avérous, 2008). Improvements in mechanical properties were obtained for the
79
blends with PLA in excess, substituting PLA for maleated PLA or by using PLA-grafted
80
amylose. Additionally, to prepare PLA/TPS blends, other authors used maleated PLA and
81
sodium montmorillonite (MMT) at 2% content, as a compatibilizing and a reinforcing
82
agent, respectively. Before processing, MMT was added either to a starch water-glycerol
83
suspension or to PLA. For the blends with 60% TPS, the location of MMT at the TPS
84
phase led to higher modulus and strength enhancement (Arroyo, Huneault, Favis, &
85
Bureau, 2010). In another report, small amounts of MMT (0.5 and 1% based on potato
86
starch dry mass) were also incorporated to a TPS/PLA blend at 60/40 composition, in a
87
multi-stage process, which consisted of potato starch gelatinization, ultrasonic dispersion
88
of MMT in water and melt mixing at 185ºC for 10 min. Reduction in water absorption and
89
improvements in mechanical and dynamic-mechanical properties were observed (Ayana,
90
Suin, & Khatua, 2014).
Ac ce p
te
d
M
an
us
cr
ip t
76
Similarly to the stabilization of foams and oil-in-water emulsions (Horozov, 2008;
91 92
Hunter, Pugh, Franks, & Jameson, 2008), nanoparticles may inhibit coalescence of
93
dispersed phase droplets in immiscible polymer blends, mainly when located at their
94
interface. However, the location of nanoparticles is dictated by thermodynamic
95
interactions, kinetic and viscosity effects (Feng, Chan, & Li, 2003; Fenouillot, Cassagnau,
96
& Majesté, 2009). Also, as postulated previously by these authors, the stabilization of
97
morphology is a requisite to maintain constant properties (Fenouillot at al., 2009). In this work, glycerol-plasticized cornstarch was melt-blended with PLA at a 70:30
98 99 100
constant composition without the addition of a compatibilizing agent, under one-step extrusion processing. Differently from the works cited before (Arroyo et al., 2010; Ayana 4
Page 4 of 32
et al., 2014), an organically-modified clay was also incorporated with increasing contents
102
(1.0 to 10.0 mass%) to the polymeric components. Both polymeric components were also
103
extruded and compression-molded alone and in the presence of the organoclay at the same
104
contents, and were characterized by X-ray diffraction and capillary rheometry. The
105
objectives of these experiments were to investigate the organoclay dispersion within TPS
106
and PLA matrices, determine changes in flow behavior resulting from the filler addition,
107
and mainly investigate the influence of these features on some properties of the resulting
108
hybrid blends. For this, TPS/PLA/organoclay blended materials were also characterized by
109
thermogravimetric analysis. The mechanical and morphological properties were
110
investigated for the extruded and compression-molded samples and for some of the
111
extruded materials , which were injection-molded .
112
Experimental
113
1.1 Materials
d
M
an
us
cr
ip t
101
Regular cornstarch (CS) was supplied by Ingredion Brasil - Ingredientes Industriais
te
114
Ltda. (São Paulo, SP, Brazil). According to the producer, this material is composed of 26–
116
30 mass% amylose and 74–70 mass% amylopectin, with less than 0.5 mass% gluten. The
117
gravimetric method was used to determine the humidity content (12 mass%). Analytical
118
grade glycerol was purchased from Vetec Química Fina (Rio de Janeiro, RJ, Brazil) and
119
was used as received. Poly(lactic acid) (PLA, NatureWorks LLC, Ingeo 2002D, with melt
120
flow index of 4-8 g/10min at 190°C / 2.16 kg) was supplied as pellets by Cargill Agrícola
121
S.A. (São Paulo, SP, Brazil). This PLA sample was previously characterized by gel
122
permeation chromatography in chloroform, and the results reported as Mn = 121.100 and
123
Mw = 197.000 (Souza, Dahmouche, Andrade, & Dias, 2013). PLA pellets were
124
cryogenically ground into powder to ensure better mixing with the other components.
125
Cloisite® 30B (C30B), a montmorillonite modified with dihydroxyethyl alkyl
Ac ce p
115
5
Page 5 of 32
126
methylammonium ions, was supplied by Southern Clay Products (Gonzales, TX, USA),
127
and was used as received.
128
1.2 Preparation of samples The mixture of CS and glycerol, added at 25 mass% based on starch dry mass, was
ip t
129
homogenized in a conventional mixer (Ika Works, Wilmington, NC, USA) for 30 min, and
131
maintained in tightly sealed polyethylene bags for 48 h at 4°C before processing to obtain
132
TPS. Glycerol-plasticized CS and PLA were mixed separately with C30B at 1.0, 2.5, 5.0,
133
7.5 and 10.0 mass% contents to obtain TPS/C30B and PLA/C30B materials. To obtain
134
TPS/PLA/C30B hybrid blends, glycerol-plasticized CS and PLA were homogenized at a
135
constant 70:30 (m/m) ratio and the organoclay was added at 1.0, 2.5, 5.0, 7.5 and 10.0
136
mass% contents, over the total mass of the polymers.
137
2.3. Extrusion and compression molding
M
an
us
cr
130
Extrusion processings were carried out in a Coperion ZSK 18 (Werner & Pfleiderer,
139
Stuttgart, Germany) co-rotating twin-screw extruder, with a L/D ratio of 40 and seven
140
heating zones. For glycerol-plasticized CS with various amounts of C30B, the seven
141
heating zones were maintained at 110, 110, 115, 115, 115, 110, 110°C, and the screw
142
speed was set at 120 rpm. For PLA and the blends (obtained by one-step extrusion
143
processings) with varying amounts of C30B, the seven heating zones were maintained at
144
160, 160, 165, 165, 165, 160, 160°C, and the screw speed was set at 120 rpm. The
145
PLA/C30B materials showed free-flowing behavior and were cooled in a water bath after
146
extrusion. All extruded materials were pelletized and compression-molded by heating at
147
different temperatures (TPS at 120ºC, PLA and the hybrid blends at 165ºC) under 68.9
148
MPa for 10 min, and cooling for 5 min in a cold press. The resulting sheets, with 12 cm x
149
10 cm x 2 cm dimensions, were used for the materials characterization.
150
2.4. X-ray diffraction (XRD)
Ac ce p
te
d
138
6
Page 6 of 32
The samples were analyzed with a Ultima IV diffractometer (Rigaku Corporation) in the
151
angular region 0.9 to 10° (2θ) at a 0.01º/s rate, in reflection mode. The CuKα1 radiation
153
with wavelength of 1.542 Å was generated at 40 kV and 20 mA. The d-values were
154
calculated by the Bragg's law (Equation 1). nλ = 2 d sen θ
155
Equation 1
ip t
152
where n is an integer, λ is the wavelength of incident wave (1.5418 Å), d is the spacing
157
between any two atomic planes in the crystal lattice, and θ is the angle of reflection.
158
2.5. Capillary viscosity measurements
us
cr
156
A Göttffert capillary rheometer model Rheograph 25 (Buchen, Germany), equipped
an
159
with a round-hole capillar of D = 1 mm (L/D = 30), was used to measure the flow
161
properties of TPS/C30B, PLA/C30B extruded materials and the hybrid blends. All tests
162
were performed over a shear rate range of 50 – 3000 s-1, at 165ºC with a prior melting
163
time of 5 min. End corrections were not applied; thus, the viscosity values are reported as
164
apparent viscosities. The setting of parameters and evaluation of raw data, as well as the
165
Rabinowitsch’s correction, were performed with the software LabRheo, provided with the
166
equipment. The consistency index and the flow behavior index were obtained according to
167
the power law model (Equations 2 and 3).
Ac ce p
te
d
M
160
168
169
τ=K
Equation 2
η=K
Equation 3
170
where τ is the shear stress, is the shear rate at the capillary wall, η is the viscosity, K is
171
the consistency index and n is the flow behavior index of the materials analyzed.
172
2.6. Thermogravimetric analysis
7
Page 7 of 32
Thermogravimetric analyses were carried out for the extruded materials under nitrogen
173
atmosphere on a TGA Q-500 equipment from TA Instruments (New Castle, USA).
175
Approximately, 10 mg of sample were heated from 25 to 700ºC, at a 10ºC/min rate.
176
2.7. Phase morphology
ip t
174
The phase morphology of the samples was examined with a Jeol electron microscope,
178
model JSM-6460LV (Akishima-shi, Japan), generally at the acceleration voltage of 20 kV.
179
Samples were cryogenically fractured before observation. The fractured surfaces were
180
vacuum-coated with gold before measurements. X-ray photoelectron spectroscopy (XPS)
181
was used to analyze the cryofractured surfaces of injection-molded samples in a
182
NORAN™ System 6 X-ray Microanalysis System, model C10015 (Thermo Fisher
183
Scientific Inc., Middleton, WI, USA), linked to the Jeol electron microscope.
184
2.8. Tensile tests of extruded and compression-molded samples
M
an
us
cr
177
Tensile tests were carried out at 21ºC in a Q-800 DMA equipment from TA Instruments
d
185
(New Castle, DE, USA), according to the ISO 527-3 method, with extruded and
187
compression-molded specimens with 13 ± 0.2 mm x 6 ± 0.2 mm x 0.8 ± 0.1 mm
188
dimensions, conditioned at 21ºC, and 50% relative humidity, for a period of at least 48 h.
189
The average value from a total of three measurements was taken. The samples were placed
190
between a fixed and a moveable clamp, and the experiments were performed in tensile
191
mode.
192
2.9. Injection molding
Ac ce p
te
186
The pelletized samples were subsequently dried in an oven at 100ºC for 2 h. Samples for
193 194
tensile tests were prepared in an injection-molding machine Arburg, model Allrounder
195
270S ( Lossburg, Germany) to obtain specimens according to ASTM D638, type I
196
(thickness 1.8 ± 0.1 mm). The five heating zones of the injection molding machine were
197
maintained at 170, 180, 190 and 200°C and 210ºC (from feed to die zone). The mold 8
Page 8 of 32
temperature was set at 30°C. The injection and the back pressures were 90 MPa and 57
199
MPa, respectively. The injection speed was 6 x 10-5 m3/s and the back time was 4 s. Before
200
performing mechanical tests, the specimens were conditioned at 21ºC and 50% relative
201
humidity for 48 h.
202
2.10. Tensile tests of extruded and injection-molded samples
ip t
198
Tensile tests were performed in a Instron Universal Testing Machine model 4204
204
(Norwood, MA, USA) equipped with a 1 kN load cell, at a speed of 1 mm/min. The
205
average value from a total of seven measurements was taken. All data were recorded and
206
processed by the software provided with the Instron equipment.
207
Results and discussion
an
us
cr
203
A diffractometer with a high resolution at low angles was used to investigate C30B
M
208
dispersion within TPS and PLA extruded materials. In Figure 1(a-c) (trace I), the 001
210
reflection for the neat C30B was observed at 2θ = 4.9º, and revealed a d001 value of 1.8 nm.
211
For TPS/C30B materials up to 5.0 mass% organoclay, no reflection was detected within
212
the 0.9 – 10º (2θ ) region (Figure 1a). This result indicated that the 001 reflection of C30B
213
was shifted to angles bellow 0.9° (2θ), with interlayer spacings higher than 9.8 nm. For the
214
TPS samples with C30B at 7.5 and 10 mass% contents, broad reflections were observed
215
around 4.5º and 4.3º (2θ), corresponding to d-values of approximately 2.0 nm. This result
216
may be attributed to the presence of a fraction of C30B aggregates, in which starch
217
molecules were partially intercalated. Although some authors reported on the
218
incompatibility of hydrophilic thermoplastic starch with C30B (Chiou et al., 2006), the
219
results of this work may be compared to those obtained by compounding potato starch with
220
C30B in an internal mixer (Park, Lee, Park, Cho, & Ha, 2003). In this latter work, no
221
reflection was found when C30B was incorporated at 2.5 mass% content, and a slight shift
222
in the organoclay 001 reflection, from around 4.71º (2θ) to around 4.30º (2θ), was
Ac ce p
te
d
209
9
Page 9 of 32
observed for the hybrids with 5 and 10 mass% C30B. The presence of two hydroxyethyl
224
moieties in the intercalant of C30B explains its interaction with starch molecules. On the
225
other hand, for all PLA/C30B materials (Figure 1b), low-intensity reflections around 2.6º
226
(2θ) were observed. The intensity of these reflections increased slightly with the increase
227
in organoclay content, and indicated that the organoclay had interlayer spacings of
228
approximately 3.4 nm, intercalated with PLA molecules. Similar results were obtained by
229
other authors for PLA and C30B (Araújo, Botelho, Oliveira, & Machado, 2014; Duan,
230
Thomas, & Huang, 2013; Fukushima, Tabuani, & Camino, 2012; Lai, Wu, Lin, & Don,
231
2014; Molinaro et al., 2013; Pluta, 2006; Zaidi et al., 2010), and attributed to hydrogen
232
bonding interaction between the hydroxyethyl groups of the oraganoclay intercalant and
233
carbonyl ester groups of PLA macromolecules. Anyway, the difference between PLA and
234
TPS intercalation might be ascribed to the bulky ring structure of TPS, because in both
235
cases hydrogen bonding interactions are favored.
d
M
an
us
cr
ip t
223
It would be worth observing the diffratograms obtained for the TPS/PLA/C30B
te
236
materials. In Figure 1c (traces II – IV), following the results for TPS/C30B, no reflection
238
was observed in the diffratograms for TPS/PLA hybrids with 1.0, 2.5 and 5.0 mass%
239
C30B, which indicated that the 001 reflection of C30B was shifted to angles bellow 0.9°
240
(2θ), with interlayer spacings higher than 9.8 nm. For the hybrids with 7.5 and 10.0 mass%
241
C30B, shallow reflections around 3º and 4.5 - 5º (2θ) were envisaged. In this case, the
242
reflection around 3º (2θ) was attributed to C30B layers intercalated with starch molecules,
243
and the other broader and less intense reflection at 4.5 – 5º (2θ) was attributed to a very
244
small amount of C30B aggregates. Anyway, these diffractograms corroborated the good
245
affinity of C30B towards both TPS and PLA phases, and suggested an even better
246
dispersion of the organoclay in the extruded blends in relation to TPS or PLA alone.
Ac ce p
237
10
Page 10 of 32
The melt flow characteristics of all extruded materials considered in this work were
248
investigated by capillary rheometry at 165ºC, and the resulting curves are shown in Figure
249
2 (a-c), as apparent viscosity versus shear rate. Shear thinning behavior was observed for
250
all samples. Comparing TPS and PLA alone, the virgin PLA presented higher viscosity
251
values, but a less pronounced shear thinning behavior compared to TPS, as indicated by the
252
higher flow index (n) value, shown in Table 1. However, for the extruded PLA sample, a
253
severe reduction of viscosity was observed. In general, processing of PLA is preceded by
254
humidity elimination, but even when this procedure was followed, degradation was
255
observed (Souza et al., 2013). Viscosity reduction of PLA was attributed to both hydrolysis
256
reaction and thermal degradation (Speranza, De Meo, & Pantani, 2014). Since the
257
objective of the extrusion processing was to produce TPS/PLA hybrid blends, in which all
258
components are hygroscopic, drying seemed an unrealistic procedure.
M
an
us
cr
ip t
247
For TPS/C30B materials (Figure 2a), the increase in organoclay content, from 1.0 to 10
d
259
mass% content, led to almost a half-decade increase of apparent viscosity, and indicated
261
that both exfoliated (at a low C30B content), and particularly intercalated organoclay
262
contributed positively to viscosity. For these materials, the flow index n (Table 1) showed
263
typical values, within the range (0.27 – 0.56) previously reported for cornstarch submitted
264
to thermomechanical treatments at different moisture contents (Vergnes & Villemaire,
265
1987). On the other hand, a substantial reduction of viscosity and n values was observed
266
for PLA/C30B materials with increasing organoclay content (Figure 2b, Table 1).
267
Recently, the thermal decomposition process of a similar system composed of PLA and
268
C30B at 0.5 and 2.5 mass% content was reported, and the nanocomposites were found
269
more stable than PLA alone (Carrasco, Pérez-Maqueda, Santana, & Maspoch, 2014). In the
270
last case, a dehumidifier was used prior to extrusion. When comparing the two hybrids, the
271
authors observed that the incorporation of 2.5 mass% C30B to PLA led to a less thermally
Ac ce p
te
260
11
Page 11 of 32
stable sample, attributed to the presence of C30B aggregates. In this study, the
273
diffractograms of Figure 1b revealed that the extrusion conditions led to intercalated C30B
274
hybrids. Thus, the reduction of viscosity, mainly at 7.5 and 10.0 mass% C30B contents,
275
and at high shear rates, might be indicating hydrolytic degradation to such an extent that
276
topological constraints were extensively minimized.
ip t
272
Viscosity curves were also obtained for the 70:30 TPS/PLA blends, with or without the
cr
277
addition of C30B (Figure 2c, Table 1). As expected, the melt viscosity for the neat
279
TPS/PLA blend resulted in intermediary values between those determined for TPS and
280
PLA, and the incorporation of C30B contributed to increase the blends viscosity. The flow
281
index values seemed to be governed by the properties of TPS, the component added at a
282
larger composition. However, no clear trend was followed by the consistency index K. It
283
might be inferred that the more pronounced Newtonian character of the TPS/PLA hybrid
284
with C30B at 10.0 mass% content (Table 1) would worsen conditions for further
285
processing, such as injection molding.
te
d
M
an
us
278
The thermal stability of TPS, PLA, and TPS/PLA blends with and without the addition
287
of C30B was evaluated by thermogravimetric analysis (TGA) under nitrogen atmosphere.
288
The TGA curves are shown in Figure 3 (a,b). Up to 100ºC, mass loss is mainly attributed
289
to humidity loss. From 100ºC to decomposition temperature, the mass loss observed for
290
TPS and TPS/PLA/C30B at 1.0 mass% might be attributed to glycerol evaporation. With
291
increasing organoclay content, a reduction of plasticizer loss was observed. The neat
292
TPS/PLA blend showed the lowest onset degradation temperature (To = 284ºC), whereas
293
TPS (To = 297ºC) and PLA (To = 332ºC) had higher thermal stabilities. For the hybrid
294
blends, To decreased moderately, from 303ºC to 288ºC, as the C30B content was increased.
295
It is worth observing in Figure 3 (c,d) (DTG curves) that the maximum degradation rate
296
was reached for PLA at Tmax = 361ºC, and for TPS at Tmax = 327ºC. While a single peak at
Ac ce p
286
12
Page 12 of 32
Tmax = 301ºC was observed for the neat TPS/PLA blend, two peaks were observed for the
298
hybrid samples. The first peak, attributed to the maximum degradation rate of the TPS
299
phase, varied from Tmax = 302ºC to Tmax = 313ºC with increasing C30B content. Lower
300
Tmax values for TPS/calcium montmorillonite hybrids in relation to TPS were previously
301
observed, and were attributed to the catalytic effect of acidic sites of montmorillonite on
302
the thermal degradation of the starch matrix (Magalhães & Andrade, 2010). Similar result
303
was also observed for other hybrid systems (Ramos Filho, Mélo, Rabello, & Silva, 2005).
304
The second peak, attributed to the maximum degradation rate of the PLA phase, varied
305
from Tmax = 357ºC to Tmax = 369ºC with increasing C30B content. This result is in
306
accordance with some authors, to whom the preparation of nanocomposites by adding
307
organo-modified montmorillonites should be considered as a promising method to reduce
308
PLA thermal degradation (Carrasco et al, 2014). In relation to PLA, other authors observed
309
a slight increase in Tmax for the PLA/C30B nanocomposites at 1.0, 3.0 and 5.0 phr
310
organoclay content (Lai et al., 2014).
te
d
M
an
us
cr
ip t
297
In this case, the TGA data also corroborated viscosity results for the TPS/PLA/C30B
312
hybrids (Figure 2), which suggested the protecting action of TPS against PLA degradation.
313
More important, DTG curves gave additional evidence on the location of clay mineral
314
particles in both TPS and PLA phases, as suggested by XRD results (Figure 1). Also, as
315
observed by some authors (Lai et al., 2014), the nucleating effect of the organoclay
316
particles should not be neglected. XRD data in the 2θ region of 2º to 35º revealed PLA
317
characteristic reflections with increasing intensities, as the content of C30B was increased
318
(Figure 1, Supplementary data).
Ac ce p
311
As in most polymer blends, TPS and PLA are immiscible. Melt blending was performed
319 320
without the addition of an organic compatibilizer and, as expected, the image of Figure 4a,
321
obtained for the neat TPS/PLA blend, shows a coarse morphology. Immiscibility was 13
Page 13 of 32
evidenced by the size, and the lack of adhesion between the two phases. For the extruded
323
and compression-molded TPS/PLA/C30B materials in Figure 4 (b-f), the micrographs
324
revealed that, as the organoclay content was increased from 2.5 to 10 mass%, at 10
325
mass% the PLA dispersed phase became undistinguished from the TPS matrix. This
326
result was related to the role of the organoclay in the improvement of the system
327
morphology. . Considering that the organoclay interacted similarly with both components,
328
and considering the viscosity results obtained, a rough estimation of viscosity ratios might
329
be envisaged for the 70:30 TPS/PLA blends with added C30B. With increasing C30B
330
contents, the viscosity ratio p and the capillary number might be expected to decrease, and
331
p might approach the value p = 1. Contrarily to the morphology improvement, previously
332
reported for C30B nanocomposites of TPS and poly(3-hydroxybutyrate-co-3-
333
hydroxyvalerate) (PHBV) (Magalhães, Dahmouche, Lopes, & Andrade, 2013), no clear
334
visualization of size reduction of the PLA dispersed phase was observed, as the content of
335
C30B was increased from 2.5 to 10 mass%. Thus, no indication on the location of the
336
organoclay nanoparticles at the interface between the components could be suggested.
337
Also, it was claimed that clay mineral particles would be incorporated preferentially into
338
the polymer with the lowest melting temperature (Fenouillot et al., 2009). In this case, TPS
339
has a lower melting temperature than PLA. However, as shown in Figure 2 for TPS/C30B
340
and PLA/C30B hybrids, with increasing contents of the organoclay, while the viscosity of
341
TPS was increased, the viscosity of PLA was decreased, and the melting pattern might be
342
changed even during the extrusion step. As the viscosity of the polymeric components
343
seems to interfere on the location of the clay minerals particles (Naderi, Lafleur, & Dubois,
344
2008), it might be presumed that C30B tended to be almost equally incorporated in both
345
TPS and PLA phases during processing, and eventually at the interface, as its content was
346
increased. However, it seems as if the amount of organoclay was not enough to saturate
Ac ce p
te
d
M
an
us
cr
ip t
322
14
Page 14 of 32
347
the interface between the TPS matrix and PLA droplets, probably because exfoliated
348
organoclay layers or organoclay aggregates were preferentially located within both
349
polymers. The mechanical properties of the TPS/PLA hybrid blends were investigated by two
351
methods, using compression-molded (Figure 5) and injection-molded specimens (Figure
352
6). These results were analyzed independently, since they were acquired by different
353
methods. In Figure 5a, for compression-molded samples, the rigid character of PLA, to
354
which the elastic modulus was determined as E = 3146.3 ± 311.8 MPa, and the soft nature
355
of TPS were once more revealed. In Figure 5b, even for the neat TPS/PLA blend, the
356
elastic modulus (190.2 ± 18.1 MPa) increased in relation to TPS alone (E = 82.3 ± 7.9
357
MPa). For the compression-molded hybrids, the moduli varied from E = 167.5 ± 14.8 MPa
358
to 501.4 ± 52.2 MPa, by incorporating 1.0 to 7.5 mass% C30B. This result was expected
359
because of the rigid nature of the filler. However, a substantial decrease in tensile stress at
360
break (σmax) was observed for the neat TPS/PLA blend (σmax = 1.5 ± 0.1 MPa) and the
361
C30B hybrids, in relation to TPS and PLA. For TPS, this value was σmax = 4.4 ± 0.4 MPa
362
and, for PLA, σmax = 5.3 ± 0.5 MPa. For the TPS/PLA hybrids, σmax varied from 1.7 ± 0.2
363
MPa for C30B incorporated at a 1.0 mass% content to σmax = 3.4 ± 0.4 MPa for the hybrid
364
with C30B at a 7.5 mass% content. As for the elongation at break, εmax, their values varied
365
for the same hybrids, respectively, from 0.8 ± 0.1% to 1.3 ± 0.01%, in relation to εmax =
366
15.2 ± 1.1% for TPS.
Ac ce p
te
d
M
an
us
cr
ip t
350
For the hybrid prepared with C30B at 10.0 mass% content, the E and σmax values
367 368
decreased to E = 383.5 ± 41.2 MPa and σmax = 2.9 ± 0.3 MPa, respectively. This result was
369
not expected, as long as increasing values were obtained with increasing contents of C30B.
370
However, the plasticizing effect of the C30B intercalant, also incorporated at a higher 15
Page 15 of 32
371
amount, might have contributed to the less rigid character of this hybrid, for which εmax
372
reached 1.9 ± 0.01%. It is worth noting that the relatively better results of tensile tests (in relation to the neat
373
TPS/PLA blend) for the hybrids, to which C30B was incorporated at 5.0 to 10.0 mass%
375
contents, followed the previously rough estimation for viscosity ratios. Because of this
376
result, these extruded samples as well as the neat 70:30 TPS/PLA blend were chosen to be
377
submitted to injection molding, to investigate possible changes of morphology and
378
mechanical properties.
us
cr
ip t
374
The mechanical properties of the injection-molded samples were evaluated (Figure 6).
380
The highest stress at break (σmax = 18.9 MPa) was obtained for the neat TPS/PLA blend.
381
Differently from the previous result obtained for the compression-molded sample, the
382
injection-molded hybrid with 10.0 mass% C30B did not show the highest ductibility. The
383
highest elastic modulus (E = 11.4 ± 3.8 MPa) and lowest elongation at break (εmax = 0.55 ±
384
0.16%) were determined for the hybrid with 7.5 mass% C30B.
te
d
M
an
379
The best ductibility was achieved for the TPS/PLA/C30B hybrid with 5.0 mass%
Ac ce p
385 386
organoclay (E = 1091.5 ± 181.3 MPa, σmax = 11.2 ± 2.2 MPa, εmax = 2.33 ± 1.4%). Bearing
387
in mind XRD results, the incorporation of 5.0 mass% C30B led to a higher degree of
388
exfoliation, which certainly contributed to the better fluidity during injection molding and
389
the highest ductibility. Moreover, the more pronounced Newtonian character of the
390
TPS/PLA hybrid with C30B at 10.0 mass%, as shown before (Table 1), indeed made
391
injection molding more difficult. To verify the effect of injection molding on morphology, cryofractured surfaces of the
392 393
injection-molded samples were analyzed by SEM (Figure 7). Although more sensitive to
394
the electron beam, the neat TPS/PLA blend in Figure 7a showed a similar image to that
395
taken for the compression-molded sample (Figure 4a). The SEM images for the hybrids 16
Page 16 of 32
with 5.0 and 7.5 mass% C30B in Figure 7 (b,c) were characterized by the appearance of
397
microvoids much larger in diameter than those observed for the corresponding
398
compression-molded samples. The appearance of these microvoids might be related to the
399
loss of some droplets of PLA dispersed phase during cryofracturing. If this supposition
400
were true, it indicated a very small amount of C30B at the interface. No particular feature
401
seems to distinguish the image observed for the injection-molded hybrid with C30B at 10
402
mass% content from that obtained for the compression-molded sample with the same
403
content of the filler. This result might be probably related to a higher amount of the
404
organoclay at the interface.
an
us
cr
ip t
396
Using XPS to analyze the cryofractured surfaces of injection-molded samples, the
406
results revealed widespread distribution of Mg, Al and Si (Supplementary data, Figures 2-
407
4), according to the previous supposition on incorporation of C30B within both
408
components phases.
409
Conclusions
te
d
M
405
The favorable interaction between the organoclay intercalant and the molecules of both
Ac ce p
410 411
components, starch and PLA, suggested by XRD and TGA results, seemed to dictate the
412
properties of extruded TPS/PLA/C30B hybrids. XRD data suggested no specific
413
preference for one of the blend components. TGA indicated the stabilizing effect of C30B
414
on the PLA phase, as the content of the organoclay increased. Shear viscosities,
415
determined for each component in the presence of the same organoclay contents, roughly
416
indicated the tendency for achieving a viscosity ratio around p = 1, as higher contents of
417
organoclay were incorporated. For the extruded and compression-molded samples, the
418
results of tensile tests followed this estimation. The SEM images revealed that the
419
morphology, even after injection molding, remained the same for the TPS/PLA hybrid
420
blend, prepared with C30B at the highest content (10.0 mass%). The high degree of 17
Page 17 of 32
organoclay dispersion, associated with the highest pseudoplasticity of the 5.0 mass%
422
hybrid blend, and the structuring characteristic of the process, certainly contributed to the
423
highest ductibility of this TPS/PLA hybrid material after processing by injection molding.
424
Acknowledgments
ip t
421
The authors acknowledge Conselho Nacional de Desenvolvimento Científico e
426
Tecnológico (CNPq), Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro
427
(FAPERJ), and Coordenação para o Aperfeiçoamento de Pessoal de Nível Superior
428
(CAPES) for financial support.
429
References
430
Araújo, A., Botelho, G., Oliveira, M., & Machado, A.V. (2014). Influence of clay organic
431
modifier on the thermal-stability of PLA based nanocomposites. Applied Clay Science,
432
88-89, 144-150.
us an
M
Arroyo, O. H., Huneault, M. A., Favis, B. D., Bureau, M. N. (2010). Processing and
d
433
cr
425
properties of PLA/thermoplastic starch/montmorillonite nanocomposites. Polymer
435
Composites, 31, 114–127.
Ac ce p
436
te
434
Ayana, B., Suin, S., & Khatua, B. B. (2014). Highly exfoliated eco-friendly thermoplastic starch (TPS)/poly (lactic acid)(PLA)/clay nanocomposites using unmodified
437
nanoclay. Carbohydrate Polymers, 110, 430–439.
438 439
Carrasco, F., Pérez-Maqueda, L. A., Santana, O.O., & Maspoch, M. Ll. (2014). Enhanced general analytical equation for the kinetics of the termal degradation of poly(lactic
440 441
acid)/montmorillonite nanocomposites driven by random scission. Polymer
442
Degradation and Stability, 101, 52-59.
443
Chiou, B.-S., Yee, E., Wood, D., Shey, J., Glenn, G., & Orts, W. (2006). Effects of
444
processing conditions on nanoclay dispersion in starch-clay nanocomposites. Cereal
445
Chemistry, 83, 300-305. 18
Page 18 of 32
446
Duan, Z., Thomas, N. L., & Huang, W. (2013). Water vapour permeability of poly(lactic acid) nanocomposites. Journal of Membrane Science, 445, 112-118.
447 448
Feng, J., Chan, C. M., & Li, J. X. (2003). A method to control the dispersion of carbon black in an immiscible polymer blend. Polymer Engineering and Science, 43, 1058–
450
1063.
Fenouillot, F., Cassagnau, P., & Majesté, J.-C. (2009). Uneven distribution of
cr
451
ip t
449
nanoparticles in immiscible fluids: morphology development in polymer blends.
453
Polymer, 50, 1333–1350.
Fukushima, K., Tabuani, D., & Camino, G. (2012). Poly(lactic acid)/clay nanocomposites:
an
454
us
452
effect of nature and content of clay on morphology, thermal and thermo-mechanical
456
properties. Materials Science and Engineering C 32, 1790-1795.
457
M
455
Horozov, T. S. (2008). Foams and foam films stabilised by solid particles. Current Opinion in Colloid & Interface Science, 13, 134–140. Huneault, M. A., Li, H. (2007). Morphology and properties of compatibilized
te
459
d
458
polylactide/thermoplastic starch blends. Polymer, 48, 270–280.
Ac ce p
460 461
Hunter, T. N., Pugh, R. J., Franks, G. V., & Jameson, G. J. (2008). The role of particles in
462
stabilising foams and emulsions. Advances in Colloid and Interface Science, 137, 57–
81.
463 464
Lai, S.-M., Wu, S.-H., Lin, G.-G, & Don, T.-M. (2014). Unusual mechanical properties of melt-blended poly(lactic acid) (PLA)/clay nanocomposites. European Polymer
465
Journal, 52, 193-206.
466 467
Luckachan, G.E., & Pillai, C.K.S. (2011). Biodegradable polymers – a review on recent
468
trends and emerging perspectives. Journal of Polymers and the Environment, 19, 637-
469
676.
19
Page 19 of 32
470
Magalhães, N. F., & Andrade, C. T. (2010). Calcium bentonite as reinforcing nanofiller for thermoplastic starch. Journal of the Brazilian Chemical Society, 21, 202-208.
471 472
Magalhães, N. F., Dahmouche, K., Lopes, G. K., & Andrade, C. T. (2013). Using an organically-modified montmorillonite to compatibilize a biodegradable blend. Applied
474
Clay Science, 72, 1–8.
ip t
473
Molinaro, S., Romero, M. C., Boaro, M., Sensidoni, A., Lagazio, C., Morris, M., & Kerry,
476
J. (2013). Effect of nanoclay-type and PLA optical purity on the characteristics of
477
PLA-based nanocomposite films. Journal of Food Engineering, 117, 113-123.
478
Naderi, G., Lafleur, P. G., & Dubois, C. (2008). The influence of matrix viscosity and
an
us
cr
475
composition on the morphology, rheology, and mechanical properties of thermoplastic
480
elastomer nanocomposites based on EPDM/PP. Polymer Composites, 29, 1301-1309. Park, H.-M., Lee, W., Park, C.-Y., Cho, W.-J., & Ha, C.-S. (2003). Environmentally
d
481
M
479
friendly hybrids. Part I Mechanical, thermal, and barrier properties of thermoplastic
483
starch/clay nanocomposites. Journal of Materials Science, 38, 909-915.
te
482
Pluta, M. (2006). Melt compounding of polylactide/organoclay: structure and properties of
485
nanocmposites. Journal of Polymer Science Part B: Polymer Physics, 44, 3392-3405.
Ac ce p
484
486
Ramos Filho, F. G., Mélo, T. J. A., Rabello, M. S., & Silva, S. M. L. (2005). Thermal
487
stability of nanocomposites based on polypropylene and bentonite. Polymer Degradation and Stability, 89, 383-392.
488 489
Schwach, E., Six, J.-L., & Avérous, L. (2008). Biodegradable blends based on starch and
490
poly(lactic acid): comparison of different strategies and estimate of compatibilization.
491
Journal of Polymers and the Environment, 16, 286–297.
20
Page 20 of 32
492
Souza, D. H. S., Dahmouche, K., Andrade, C. T., & Dias, M. L. (2013). Synthetic
493
organofluoromica/poly(lactic acid) nanocomposites: structure, rheological and thermal
494
properties. Applied Clay Science, 80-81, 259-266. Speranza, V., De Meo, A., & Pantani, R. (2014).Thermal and hydrolytic kinetics of PLA in the molten state. Polymer Degradation and Stability, 100, 37-41.
497
cr
496
ip t
495
Sundararaj, U., & C. W. Macosko, C. W. (1995). Drop breakup and coalescence in polymer blends: the effects of concentration and compatibilization. Macromolecules,
499
28, 2647-2657.
Vergnes, B., & Villemaire, J. P. (1987). Rheological behaviour of low moisture molten
an
500
maize starch. Rheological Acta, 26, 570-576.
501
Wang, N., Yu, J., & Ma, X. (2007). Preparation and characterization of thermoplastic
M
502
us
498
starch/PLA blends by one‐step reactive extrusion. Polymer International, 56, 1440–
Ac ce p
te
d
503
1447.
504 505
Wu, S. (1987). Formation of dispersed phase in incompatible polymer: interfacial and rheological effects. Polymer Engineering and Science, 27, 335-343.
506 507
Zaidi, L., Bruzaud, S., Bourmaud, A., Médéric, P., Kaci, M., & Grohens, Y. (2010). Journal of Applied Polymer Science, 116, 1357-1365.
508 509
Zhang, J. - F., & Sun, X. (2004). Mechanical properties of poly(lactic acid)/starch composites compatibilized by maleic anhydride. Biomacromoleules, 5, 1446-1451.
510 511 512
21
Page 21 of 32
513 514 515
ip t
516 517
cr
518
us
519 520
an
521 522
M
523 524
d
525
te
526
Ac ce p
527 528 529
Table 1: Power law constants for glycerol-plasticized thermoplastic starch, poly(acid acid)
530
and their blends, with or without the incorporation of the C30B.
531
K (Pa.s)
n
R2
TPS
10 399
0.27
0.998
TPS/ C30B (1.0 mass%)
9 727
0.31
0.999
TPS/ C30B (2.5 mass%)
9 332
0.36
0.999
TPS/ C30B (5.0 mass%)
13 489
0.31
0.999
Blend composition
22
Page 22 of 32
21 281
0.28
0.998
TPS/C30B (10.0 mass%)
25 577
0.31
0.999
PLA (as received)
44 668
0.40
0.995
964
0.32
0.996
PLA/C30B (1.0 mass%)
100 025
0.17
PLA/C30B (2.5 mass%)
51 286
0.25
PLA/C30B (5.0 mass%)
41 686
0.15
PLA/C30B (7.5 mass%)
125 892
PLA/C30B (10.0 mass%)
25 150
TPS/PLA
0.998
us
cr
0.995 0.996
0.999
-0.04
0.998
13 423
0.37
0.998
TPS/PLA/C30B (1.0 mass%)
25 118
0.29
0.997
TPS/PLA/C30B (2.5 mass%)
40 738
0.34
0.994
81 234
0.28
0.999
51 121
0.39
0.996
21 579
0.60
0.998
an
-0.44
M
PLA (after extrusion)
ip t
TPS/C30B (7.5 mass%)
te
TPS/PLA/C30B (7.5 mass%)
d
TPS/PLA/C30B (5.0 mass%)
TPS/PLA/C30B (10.0 mass%)
Ac ce p
532 533 534
Caption to the figures
535
Figure 1: X-ray diffractograms in the 2θ region of 0.6º to 10º for the neat C30B (traces I);
536
(a) TPS, (b) PLA, (c) 70:30 TPS/PLA blend with the incorporation of C30B at 1.0 (traces II), 2.5 (traces III), 5.0 (traces IV), 7.5 (traces V), 10.0 (traces VI) mass%.
537 538
Figure 2: Variation of apparent viscosity as a function of shear rate at 165ºC for
539
(a) TPS alone (■) and hybrids, (b) PLA alone (■) and hybrids, and (c) neat 70:30
540
TPS/PLA blend (■) and hybrids, with the incorporation of C30B at 1.0 (U), 2.5 ( ),
541
5.0 ({), 7.5 (V), and 10.0 () mass%.
23
Page 23 of 32
Figure 3: TGA curves for (a) TPS (¯) and PLA ( ), and (b) neat 70:30 TPS/PLA blend
543
(■) and hybrids, (c) DTG curves for TPS (¯) and PLA ( ), and (d) neat 70:30
544
TPS/PLA blend (■) and hybrids with the incorporation of C30B at 1.0% ( ), 2.5%
545
(U), 5.0 ({) 7.5 % (V), and 10.0 () mass%.
ip t
542
Figure 4: SEM images for the extruded samples; neat 70:30 TPS/PLA blend (a), and for
547
the hybrids with the incorporation of C30B at 1.0 (b), 2.5 (c), 5.0 (d), 7.5 (e) and 10.0
548
(f) mass%.
us
cr
546
Figure 5: Stress versus strain curves for the extruded samples; (a) TPS ( ) and PLA (¯);
550
(b) neat 70:30 TPS/PLA blend (■) and hybrids with the incorporation of C30B at 1.0%
551
( ), 2.5% (U), 5.0 ({) 7.5 % (V), and 10.0 () mass%.
an
549
Figure 6: Stress versus strain curves for the injection-molded samples; neat 70:30
553
TPS/PLA blend (■) and for the hybrids the incorporation of C30B at 5.0 ({) 7.5 %
554
(V), and 10.0 () mass%.
d
Figure 7: SEM images for the injection-molded samples; neat 70:30 TPS/PLA blend (a),
te
555
M
552
and for the hybrids the incorporation of C30B at 5.0 (b) 7.5 % (c), and 10.0 (d) mass%.
Ac ce p
556 557 558 559
24
Page 24 of 32
ip t cr M
an
us
560
Ac ce p
te
d
561
562 563 564
565 566 567 568
Figure 1
569
25
Page 25 of 32
cr
ip t
570
M
an
us
571
Ac ce p
te
d
572
573 574 575 576 577
Figure 2
26
Page 26 of 32
ip t M
an
us
cr
578 579
Ac ce p
te
d
580 581
582 583
584 585
Figure 3
27
Page 27 of 32
ip t cr M
an
us
586
Ac ce p
te
d
587
588
589 28
Page 28 of 32
590 591 592
ip t
Figure 4
an
us
cr
593
Ac ce p
te
d
M
594
595 596 597 598 599
600 601 602 603
29
Page 29 of 32
604 605
Figure 5
606
ip t
607
M
an
us
cr
608
609 610
te
d
611 612
Ac ce p
613 614 615 616 617 618 619 620 621 622 623
30
Page 30 of 32
Figure 6
cr
ip t
624
d
M
an
us
625
Ac ce p
te
626
627 628
629 fig. 7
31
Page 31 of 32
638
Highlights
639 640
• Melt-blending of cornstarch hybrid blends were investigated.
ip t
641
643
cr
642
• Hybrids of both components were prepared at the same contents of the organoclay.
645
us
644
• Results indicated the incorporation of the organoclay in both phases.
647
an
646
• .
• The organoclay dispersion and pseudoplasticity affected hybrid materials properties.
d
649
M
648
te
650 651
Ac ce p
652 653 654 655 656 657 658 659
660
33
Page 32 of 32
