Composites Part A 124 (2019) 105472

Contents lists available at ScienceDirect

Composites Part A
journal homepage: www.elsevier.com/locate/compositesa

Impact of the natural filler babassu on the processing and properties of T

PBAT/PHB films
R. Hoffmanna,b, D.D.S. Moraisc, C.J.F. Brazc, K. Haaga, R.M.R. Wellena,c,d, E.L. Canedoc,
⁎
L.H. de Carvalhoc, K. Koscheka,
a
Fraunhofer Institute for Manufacturing Technology and Advanced Materials (IFAM), Wiener Str. 12, 28359 Bremen, Germany
b
University of Bremen, Department 2 Biology/Chemistry, Leobener Str. 3, 28359 Bremen, Germany
c
Federal University of Campina Grande, Department of Materials Engineering, Campina Grande 58249-140, Brazil
d
Federal University of Paraíba, Department of Materials Engineering, João Pessoa 58051-085, Brazil

A R T I C LE I N FO A B S T R A C T

Keywords: Melt extruded films consisting of polybutyrate adipate terephthalate (PBAT), poly(3-hydroxybutyrate) (PHB)
A. Biocomposite and babassu as a natural and cost-effective filler were studied aiming at the preparation of partly bio-based and
B. Mechanical properties biodegradable films. PHB suffers from low thermal stability entailing a narrow processing window. Rheological,
B. Thermal properties thermal and mechanical properties revealed that an effective processing of the PBAT/PHB blends with well-
E. Extrusion
dispersed babassu is possible. Thermal degradation of the composites during processing and the mechanical
properties are driven mainly by the polymer composition and less by the filler. However, babassu addition
reduces the crystallization rate and facilitates by this the handling of the extruded films prior to solidification.
Among the composites investigated, the composites based on 50%PBAT/50%PHB blends combined both, best
processing and handling parameters as well as thermal and mechanical material properties. Babassu was found
to facilitate PHB processing while maintaining the overall mechanical properties and giving access to ecologi-
cally friendly, cost-effective composite materials.

1. Introduction mechanical properties to polyolefins such as polypropylene (PP) [12],

its widespread application is limited because of its high cost and certain
Nowadays, large quantities of plastics are widely used, for instance, inferior material properties such as brittleness and low impact re-
in automotive, packaging, health, and electric-electronic industries to sistance [13]. Moreover, PHB suffers from low thermal stability en-
fulfill modern society needs [1]. However, these materials have dele- tailing a narrow processing window. The degradation effect of con-ro-
terious effects on the environment as most of the plastic based products, tational mixing and extruding was studied in detail and is well
particularly for packaging applications, are very rapidly disposed. Thus, described in the literature, pointing chain scission as consequence of
the amount of plastic waste in our environment grows at an astounding thermal and thermal-oxidative degradation [14,15]. For extrusion of
rate threatening our ecosystem [2]. Biodegradable polymers that can be PHB, Pachekoski et al. reported a decrease of the average molar mass
converted into carbon dioxide by microorganisms represent an alter- via size exclusion chromatography by 29% and a correlated increase of
native to conventional polymers [3–6]. the melt flow index [16].
PBAT is a biodegradable random copolymer based on fossil re- Blending with other biopolymers such as PLA [17], PCL [18] or
sources made up of adipic acid, 1,4-butanediol and dimethyl ter- other additives, such as tannic acid as a natural phenol based additive
ephthalate. With an elongation at break up to 700% PBAT is very [19] represents a major approach to adjust thermal properties of PHB.
ductile [7,8] but in its mechanical properties it resembles low density Depending on the process, blending of PHB with PLA requires, how-
polyethylene (LDPE). Among others, it is used for compostable bio- ever, synthetic petrochemical additives, such as Joncryl®, in order to
waste bags (according to EN 13432) or in soil biodegradable mulch overcome compatibility issues, to achieve sufficient mechanical stabi-
films (according to EN 17033) [3,9–11]. lity and to reduce thermal degradation [20–22]. Alternative routes for
PHB is a biodegradable semi-crystalline polymer and originates tailoring PHB by incorporating additional functional groups via che-
from bacterial fermentation of sugar cane. Although PHB has similar mical modification alter its properties inherently [23].

⁎
Corresponding author.
E-mail address: [email protected] (K. Koschek).

https://doi.org/10.1016/j.compositesa.2019.105472
Received 20 November 2018; Received in revised form 29 May 2019; Accepted 30 May 2019
Available online 31 May 2019
1359-835X/ © 2019 Elsevier Ltd. All rights reserved.
R. Hoffmann, et al. Composites Part A 124 (2019) 105472

Beside the impact on thermal properties, blending of PHB with other Table 1
polymers affects physico-chemical influencing the overall materials Molar masses of PHB and PBAT.
performance. Epoxy and styrene functionalized acrylates were de- Compound Mn (kDa) Mw (kDa) Polydispersity
scribed to affect the biodegradability [24], as the addition of hydro-
phobic long chain polymers and aromatic esters lower the biode- PHB 192.028 273.204 1.42
PBAT 51.79 112.66 2.17
gradation [25].
The addition of i.e. PBAT [26] or aliphatic-butylene terephthalate
copolyester [27] diminishes PHB’s crystallizability. Beber et al. [26]
2.2. Methods1
suggested that PBAT macromolecules at the front or around the
growing crystallites of PHB would be slowing down and even pre-
2.2.1. Specimen processing
venting the crystal growth of PHB. Whereas, Wojtczak et al. [27] dis-
PBAT, PHB and babassu were thoroughly dried (20 h/60 °C under
cussed that primary nuclei in PHB phase migrate to the copolyester
350 mbar) and mixed for 10 min in a HAAKE Rheomix QC Lab Mixer for
phase during compounding.
the HAAKE PolyLab QC System (ThermoFisher, USA) fitted with high
Aiming to improve PHBs flexibility for potential packaging appli-
intensity (roller type) rotors and infrared torque measurement, oper-
cations, Modi et al. [28] synthesized PHB with various co-polymers
ated at a nominal speed of 60 rpm while maintaining the chamber wall
such as poly-(3-hydroxyvalerate) (HV) and observed a decrease of the
at a constant temperature (160 °C).
glass transitions and melting temperatures. In addition, HV incorpora-
Three PBAT/PHB blends containing 25%, 50% and 75% of each
tion broadened the processing window since it improved melt stability
component were tested. Compounds containing 1 wt% and 3 wt% ba-
at lower processing temperatures. The complex viscosity decreased
bassu filler, with each of the three blends as matrices, were also pre-
with increasing temperature due to a decrease in molecular weight of
pared. Compositions and codification are presented in Table 2. All
the samples. According to the authors, 160 °C is the optimal tempera-
components were premixed and fed together into the internal mixer
ture to process the copolymers.
(“time 0”). Total mass was always 275 g, which – taking into con-
Biodegradable polymers such as PHB and PLA are still expensive.
sideration that for both polymers density is about 1.25 g/cm3 [37] and
Therefore, they are compounded with other polymers as well as fillers
the free volume of the processing chamber is 310 cm3 – corresponds to a
aiming at cost reductions along with the improvement of material
fill factor of (approximately) 70% at ambient temperature.
properties [29–32]. To circumvent the use of petrochemical products,
The output of the mixer was ground, dried for 48 h at 60 °C under
bio-based starch or cellulose-based fillers can be co-compounded. Ba-
350 mbar, and processed in a 16-mm laboratory single-screw extruder
bassu for example represents natural and cost-effective filler originating
(AX Plasticos, Brazil), operated at 170 °C and 28 rpm, fitted with a flat
naturally from a palm tree found in the North, Northeast and Midwest
film die and roll attachment to obtain flat films of approximately
regions of Brazil. Usually its shell is discarded and burned without any
100 μm thickness.
benefit. In polymer composites babassu yields both, material and
technology related improvements as well as economic benefits [33].
As neat PHB features challenges such as brittleness, limited cost-
2.2.2. Thermogravimetric analysis (TGA)
effectiveness and thermal degradation while processing, this contribu-
TGA experiments were performed with a DTG 60H apparatus
tion aims at developing of PHB based films with improved properties
(Shimadzu, Japan). Scans were acquired from room temperature
suitable for potential packaging and mulch films for food and agri-
(≈23 °C) to 650 °C with a heating rate of 10 K/min and a nitrogen flow
culture applications. This will be accomplished by blending PHB with
of 100 mL/min. Evaluation of the TGA data was performed according to
PBAT as a plasticizer and compounding with babassu mesocarp as
DIN EN ISO 11358-1:2014.
natural and cost-effective filler. Blending PHB with PBAT is subject to
challenges in miscibility at low processing temperatures [34], that were
overcome by the choice of processing conditions and a two-step treat-
2.2.3. Differential scanning calorimetry (DSC) measurements
ment. Thermal and mechanical analysis will be used to explain pro-
DSC tests were performed with 7–8 mg samples of films in a DSC
cessing related observations.
Q20 equipment (TA Instruments, USA) under a nitrogen flow of 50 mL/
min. Samples at ambient temperature (≈23 °C) were heated to 250 °C,
cooled to 0 °C, and reheated to 250 °C; heating and cooling rates were
2. Experimental
kept at 10 K/min.

2.1. Materials
2.2.4. Tensile experiments
Poly(butylene adipate-co-terephthalate) (PBAT), was provided by
Specimens for tensile tests were cut perpendicular to extrusion di-
BASF as Ecoflex® F Blend C1200, and poly(3-hydoxybutyrate) supplied
rection with approximately 100 × 10 mm2. The exact dimensions of
by PHB Industrial SA (Brazil) and commercially available as Biocycle®
each specimen were determined at three points for length, width and
1000. Both polymers were used without any further treatment.
thickness. Average values of width and thickness were used as input
Table 1 represents the molar masses of the as received samples of
data for strength and stiffness calculations during mechanical testing
PHB and PBAT, taken from measurements of Nunes et al. [24] and
experiments, which were carried out on a DL-10000 tensile testing
Montoro et al. [35], respectively.
machine (EMIC, Brazil) operating with a 200 kgf load cell. For the
Mesocarp of the babassu nut shell obtained from babassu palm tree
compounds with 25% and 50% PBAT the extension rate was 5 mm/min
(Orbignya phalerata sp, Mart. (Aracaceae)) was received as micronized
whereas for the compounds with 75% PBAT an extension rate of
babassu mesocarp supplied by Florestas Brasileiras S.A. (Brazil), and
50 mm/min was used. All tests were performed at room temperature
screened (#200 to #325 standard ASTM screens). The screened babassu
(≈23 °C) and ambient humidity. 10 specimens of each composition
powder was used as received, without pre-processing. Babassu meso-
were tested, and the average values are reported.
carp is composed of cellulose (45%), hemicellulose (34%), and lignin
(18%) with 3% of mineral ashes [36].

1
A detailed theoretical background concerned with rheological and thermal
characterizations is provided in Appendix A1 at the end of this manuscript.

2
R. Hoffmann, et al. Composites Part A 124 (2019) 105472

Table 2
Samples composition and codes.
Code Mass (g)

PBAT PHB babassu Total

25%PBAT/75%PHB 68.75 206.25 0.00 275.00

50%PBAT/50%PHB 137.50 137.50 0.00 275.00
75%PBAT/25%PHB 206.25 68.75 0.00 275.00
25%PBAT/75%PHB – 1%babassu 68.06 204.19 2.75 275.00
50%PBAT/50%PHB – 1%babassu 136.13 136.13 2.75 275.00
75%PBAT/25%PHB – 1%babassu 204.19 68.06 2.75 275.00
25%PBAT/75%PHB – 3%babassu 66.69 200.06 8.25 275.00
50%PBAT/50%PHB – 3%babassu 133.38 133.38 8.25 275.00
75%PBAT/25%PHB – 3%babassu 200.06 66.69 8.25 275.00

3. Results and discussion

3.1. Degradation during rheomixing

Blending was performed by rheomixing in an internal mixer

yielding well-homogenized blends and composites of PBAT, PHB and
babassu. The appropriate films were obtained successfully by flat-die Fig. 2. Torque adjusted to 160 °C as a function of time during the last two
extrusion. Based on the literature database and in previous works of minutes of processing, for the PBAT/PHB blends – 1% babassu. (For inter-
Modi et al. [28], rheomixing was performed at 160 °C as test tem- pretation of the references to colour in this figure legend, the reader is referred
perature and for 10 min compounding time, respectively, in order to to the web version of this article.)
prevent degradation in PBAT/PHB/babassu blends and composites.
Oppositely to conventional blending, rheomixing allows to measure viscous dissipation is engaged. Polymers are substantially molten after
torque, temperature and total energy feed during processing. This leads 5–6 min processing time. No steady state plateau was reached, neither
to the in-process determination of possible degradation of the material by temperature nor by torque, within the 10 min of processing.
at given conditions as well as the change in viscosity due to homo- Torque depends on polymer viscosity (and hence molecular weight),
genization and possible chain scission processes. which strongly depends on temperature. Torque will decrease if poly-
The raw data, i.e. temperature and torque versus time during pro- mers degrade during processing but it will also decrease if temperature
cessing of the compounds studied in this work is shown in Fig. 1 ex- is raised. In order to eliminate temperature dependence on torque, the
emplarily for samples with 1 wt% babassu. (Additional data and cal- concept of adjusted torque (i.e., the torque expected for a given mate-
culations can be found in Supplementary Material.) Temperature rial to achieve at a fixed reference temperature) is used. A detailed
initially drops as cold material is fed into the mixer and then increases description on the fundamentals of this concept can be found elsewhere
by the combined effect of mechanical energy dissipation and heat [36]. Torque was adjusted at the reference temperature T* = 160 °C
transfer to/from the chamber wall. Torque increases by mechanical according to Eq. (3) (see Appendix A), using β = 0.02 °C−1. Fig. 2
energy dissipation in the solids mix (friction and plastic deformation of shows the results for the last two minutes of processing (8–10 min).
the polymer particles), then decreases as polymers melt. Subsequently A higher torque was observed for PBAT-rich samples and a slight
the solid dissipation mechanisms are switched off and the less efficient decrease in torque with time for all samples. This suggests higher

Fig. 1. Evolution of temperature (left) and torque (right) during processing of the PBAT/PHB/1%babassu compounds in the laboratory internal mixer. (For inter-
pretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3
R. Hoffmann, et al. Composites Part A 124 (2019) 105472

25%PBAT/75%PHB 100 75%PBAT/25%PHB

20 50%PBAT/50%PHB
75%PBAT/25%PHB - 1% Babassu
75%PBAT/25%PHB - 3% Babassu
75%PBAT/25%PHB 50%PBAT/50%PHB
80 50%PBAT/50%PHB - 1% Babassu
50%PBAT/50%PHB - 3% Babassu
15 25%PBAT/75%PHB
Z* (Nm)
25%PBAT/75%PHB - 1% Babassu
60 25%PBAT/75%PHB - 3% Babassu

Mass (%)
10
40

5 20

0 0
0 1 3 250 300 350 400 450 500 550 600
babassu content (%) Temperature (°C)

Fig. 3. Mean temperature adjusted torque Z* during the last two minutes of Fig. 5. TGA plots of PBAT/PHB blends and PBAT/PHB/babassu composites.
processing for PBAT/PHB blends and PBAT/PHB/babassu compounds. (For (For interpretation of the references to colour in this figure legend, the reader is
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
referred to the web version of this article.)
3.2. Thermal stability of extruded films
relative torque drop rate (min-1)

TGA experiments were performed to elucidate the effect of com-

0.08 25%PBAT/75%PHB
50%PBAT/50%PHB pounding on PBAT/PHB/babassu’s thermal stability. Fig. 5 displays the
75%PBAT/25%PHB acquired thermograms, whereas detailed numeric results can be found
0.06 in the Supplementary Material. No significant weight loss (degradation)
is observed in TGA up to 200 °C, despite evaporating water. The de-
gradation observed during rheomixing is due to chain scission of
0.04
polymeric chains to rather shorter chains than to volatile compounds as
confirmed by TGA experiments.
0.02 Above 200 °C, all blends and composites start decomposing in three
distinct steps. The first step of weight loss for the neat blends with the
initial temperature of 260 °C was attributed to PHB’s degradation, as
0.00
0 1 3 the weight loss changes fits to the mass ratio of the prepared material
babassu content (%) (weight loss of 73% for the blend 25%PBAT/75%PHB). The second step
in weight loss corresponds mainly to the degradation of the PBAT phase
Fig. 4. Relative rate of torque drop during the last two minutes of processing with an initial temperature around 360 °C. Degradation temperatures of
(rate of degradation) for PBAT/PHB/babassu compounds. (For interpretation of PHB and PBAT determined in TGA experiments of the neat polymers
the references to colour in this figure legend, the reader is referred to the web corresponded to the ones that were determined in the blends con-
version of this article.)
firming a correct assignment of the observed degradation steps in the
blends. Furthermore, the ratio of the single degradation steps equaled
degradation rates for the PHB matrix rich blends and composites. the initial PHB/PBAT ratio confirming the homogeneity of the batches
Torque decrease at the final processing stages indicates chain scission, that is also represented in the microscopic amount of sample taken for
which results in lower viscosity and thus allows for a less energy con- TGA.
suming mixing process. The relationship of polymer degradation rate Upon addition of 1 and 3 wt% babassu, the initial PHB degradation
and viscosity determined by torque rheometry is demonstrated in the temperatures in the blends decreases to 219 °C (25%PBAT/75%PHB),
Appendix A according to Wetzel and Shih [15]. To quantify these 221 °C (50%PBAT/50%PHB), and 251 °C (75%PBAT/25%PHB), re-
trends, the mean adjusted torque was computed, as shown in Fig. 3, and spectively. A similar decrease in thermal stability was observed for
the relative rate of decrease estimated with Eq. (5) as shown in Fig. 4. PBAT’s degradation step. This indicates a reduced thermal stability
Numerical results can be found in Supplementary Material. caused by the presence of babassu.
Fig. 3 shows the results for temperature-adjusted torque of all in- Tangential analysis of the decomposition steps revealed for the ex-
vestigated compounds. Torque is fairly insensitive to babassu content trapolated starting temperature TA and end temperature TC a decrease
within the ranges investigated (1–3 wt%), if anything, results suggest a upon addition of babassu for the first decomposition step. However,
discrete reduction upon the addition of filler (1 wt% or 3 wt% babassu there is no clear tendency in the deviation of the tangential step tem-
mesocarp). This finding perhaps could be attributed to enhanced matrix perature, which sustains the degradation of PHB and PBAT is in-
degradation by the filler, enough to overcome the expected slight in- dependently of the babassu concentration.
crease in viscosity due to the presence of a solid load. At the low con- Against this background, the weight loss trend is clearly identified –
centrations tested, this effect was more pronounced for PHB. the higher the PBAT content, the more thermally stable the blend is,
Fig. 4 shows a very significant increase of the rate of degradation whereas babassu addition slightly decreases the thermal stability of all
with increasing PHB content in the matrix, indicating that the PHB PBAT/PHB/babassu composites. Changes of thermal stability based on
degrades much faster than PBAT, in accordance with our previous ex- babassu are only observed at the first degradation step and not for the
perience [36]. The dependence of the rate of matrix degradation on second decomposition at elevated temperatures [17,18]. Basically, TGA
filler content is not clear. At any case, filler incorporation did not seem results are in agreement with the rheological characterization.
to increase the rate of degradation during the final processing stage.

3.3. Effect of blending on thermal properties

Thermal properties of the PBAT/PHB blends and the appropriate

4
R. Hoffmann, et al. Composites Part A 124 (2019) 105472

Fig. 6. Typical DSC output plot for a 50%PBAT/50%PHB blend. Blue: heat flows (left scale), red: temperature (right scale). (For interpretation of the references to
colour in this figure legend, the reader is referred to the web version of this article.)

babassu compounds were investigated by DSC. Fig. 6 shows a typical C1 75%PBAT/25%PHB

1.2
crystallization rate (min-1)
DSC output of 50%PBAT/50%PHB blend showing four phase-changing 75%PBAT/25%PHB - 1% babassu
75%PBAT/25%PHB - 3% babassu
events: melting during first heating stage (coded F1), crystallization 50%PBAT/50%PHB
50%PBAT/50%PHB - 1% babassu
50%PBAT/50%PHB - 3% babassu
from the melt in the cooling stage (coded C1), cold crystallization upon 0.9 25%PBAT/75%PHB

reheating (coded C2), and finally a second melting event during the 25%PBAT/75%PHB - 1% babassu
25%PBAT/75%PHB - 3% babassu

second heating (coded F2). In this work the melt and cold crystal-
lization events (C1 and C2) and the second melting event (F2) of the 0.6
blends and compounds investigated were considered and will be dis-
cussed in the following. 0.3

3.3.1. Melt crystallization events, C1 0.0

DSC scans for the melt crystallization of the compounds investigated 80 60 40
in this work are presented in Fig. 7. Kinetic and thermodynamic para- temperature (°C)
meters computed from these scans can be found in the Supplementary
Fig. 8. Melt crystallization rate of PBAT/PHB blends and PBAT/PHB/babassu
Material. DSC results indicate a single melt crystallization event for all
composites. (For interpretation of the references to colour in this figure legend,
blends and composites in the temperature range of 60–80 °C. Solely the the reader is referred to the web version of this article.)
25%PBAT/75%PHB blend presented melt crystallization peaks of PBAT
and PHB as distinct events, whereas the peak at higher temperature of
decreases down to 0.56 min−1 and 0.61 min−1 upon addition of 1 and
125 °C was attributed to PHB and the one at lower temperature to
3 wt% babassu, respectively. The melt crystallization rate decreases for
PBAT. The melt crystallization rates for the compounds investigated in
all compounds when babassu is added to the blends. The reduced
this work are depicted in Fig. 8, showing the highest crystallization rate
crystallization rate affected the extrusion processing of babassu com-
for the blend 50%PBAT/50%PHB with a rate of 1.35 min−1 which
posite films. The films were easier to handle and stayed longer flexible
1.0 to be convoluted on a film roll compared to films of the neat blends.
C1 75%PBAT/25%PHB Plots of relative crystallinity as a function of temperature deducted
75%PBAT/25%PHB - 1% babassu
75%PBAT/25%PHB - 3% babassu
50%PBAT/50%PHB
from Fig. 8 are presented in Supplementary Material. Sigmoidal curves
50%PBAT/50%PHB - 1% babassu
50%PBAT/50%PHB - 3% babassu
without discontinuities which are characteristic for phase transitions in
heat flow (mW)

0.8 25%PBAT/75%PHB
25%PBAT/75%PHB - 1% babassu polymers were obtained for all compounds. Similar behavior was pre-
25%PBAT/75%PHB - 3% babassu
viously reported by Ries et al. [38] in PHB/ZnO composites and by Feng
et al. [39] in biocomposites with PBAT.
0.6

3.3.2. Cold crystallization events, C2

DSC scans for the cold crystallization of the PBAT/PHB blends and
0.4 their corresponding babassu composites are presented in Fig. 9. Basi-
140 120 100 80 60 40 cally, it is known that PHB cold crystallizes [40], whereas, the amount
temperature (°C) of cold crystallization is higher, the higher the cooling rate is, thus, less
time is available for macromolecular rearrangement entailing lower
Fig. 7. DSC scans of melt crystallization of PBAT/PHB blends and PBAT/PHB/ melt crystallization.
babassu composites. (For interpretation of the references to colour in this figure For the neat 25%PBAT/75%PHB blend cold crystallization occurs to
legend, the reader is referred to the web version of this article.) a low degree as most of the PHB and PBAT melt crystallized. Based on

5
R. Hoffmann, et al. Composites Part A 124 (2019) 105472

75%PBAT/25%PHB
0.9 75%PBAT/25%PHB - 1% babassu C2
75%PBAT/25%PHB - 3% babassu -0.4
50%PBAT/50%PHB
50%PBAT/50%PHB - 1% babassu

0.6
50%PBAT/50%PHB - 3% babassu F2

heat flow (mW)

25%PBAT/75%PHB
-0.6
heat flow (mW)
25%PBAT/75%PHB - 1% babassu
25%PBAT/75%PHB - 3% babassu

0.3
-0.8
75%PBAT/25%PHB
75%PBAT/25%PHB - 1% babassu
0.0 75%PBAT/25%PHB - 3% babassu
50%PBAT/50%PHB
-1.0 50%PBAT/50%PHB - 1% babassu
50%PBAT/50%PHB - 3% babassu
25%PBAT/75%PHB
-0.3 25%PBAT/75%PHB - 1% babassu
25%PBAT/75%PHB - 3% babassu

20 40 60 80 -1.2
120 140 160
temperature (°C) temperature (°C)
Fig. 9. DSC scans of cold crystallization of PBAT/PHB blends and PBAT/PHB/ Fig. 10. DSC scans of second melting of PBAT/PHB blends and PBAT/PHB/
babassu composites during heating. (For interpretation of the references to babassu composites. (For interpretation of the references to colour in this figure
colour in this figure legend, the reader is referred to the web version of this legend, the reader is referred to the web version of this article.)
article.)

composites, the degree of cold crystallinity increased with PHB content

the results that PBAT and babassu suppressed melt crystallization of
and no significant change was observed upon compounding with ba-
PHB, the observed cold crystallization events of blends with higher
bassu. Only slight changes were observed for the crystallinity developed
PBAT content and babassu composites were attributed to PHB as main
during melt crystallization and second melting. In applications where
parts of PBAT melt crystallized (C1) beforehand.
low crystallinity is required, a high PBAT content is suggested.
Regarding the crystallization exotherms PBAT seems to hinder PHB
Additionally, high reheating rates can be applied which would hinder
melt crystallization, which therefore mainly cold crystallizes after
the macromolecular reorganization and thus decrease the PHB crys-
PBAT’s melt crystallization. In other words, PHB crystallizes during
talline phase. These results are in agreement with the research of Costa
heating in contact with semi crystalline solid PBAT and solid babassu
et al. who reported a dependence of the crystallinity on the cooling rate
particles. This is in agreement with the studies performed by Beber
in PBAT/PHB blends [42]. Attention should be given to the desired
et al. [26]. Such crystallization behavior could lead to the formation of
properties, since varying the composition will have a direct influence
a morphological microstructure with crystals exhibiting miscellaneous
on the compound’s mechanical properties.
thermal stability, i.e., diversified crystallinity and perfection. This
would explain the main driving force for the complex melting behavior
discussed above. From the scans in Fig. 9 cold crystallization rates were 3.4. Mechanical properties of PBAT/PHB/BABASSU compounds
computed indicating that compounds with higher PHB content crys-
tallized at lower temperature around 48 °C, i.e., longer times and wider Tensile tests of PBAT/PHB/babassu films were performed to analyze
undercooling (ΔS = Tmo − Tc ). Increasing the required undercooling their mechanical resistivity (Fig. 11). As the mechanical performance is
and/or decreasing the crystallization rate will have a significant impact affected by the sample composition and morphology, equal polymer
on the operating processing window that is affected by the crystallinity compositions and homogeneity were confirmed by thermal analysis.
of the material. However, in a temperature range as low as 30–60 °C the Moreover, the degree of babassu dispersion and morphology was shown
driving forces for macromolecular rearrangement are ceasing, i.e., the to be homogeneous among the entire sample with SEM imaging in a
polymers are practically in the solid state. That makes melt crystal- comprehensive study published by Costa et al. [34].
lization of PHB difficult and promotes therefore the cold crystallization An increase in Younǵs modulus from 326 MPa to 536 MPa was ob-
during subsequent heating. This morphological state might be con- served for the blends changing from 25 to 50% PHB, and up to 557 MPa
sidered as a key control variable of the resulting microstructure and by in case of the blend composition 25%PBAT/75%PHB. As expected,
that the materials properties. compounds with higher PHB content presented higher Young’s modulus
Relative crystallinity for the cold crystallization was calculated from compared to PBAT rich compounds. However, changing from 25 to
Fig. 9. All plots have a sigmoidal form associated to crystallization 50% PHB caused a disproportionate step in Younǵs modulus of 39% in
without discontinuity, following a similar trend observed in melt comparison to the one changing from 50 to 75% PHB that is marginal.
crystallization. In contrast, PBAT in blends and composites affects positively the tensile
strength and elongation at break. Analog to the effect on stiffness, the
3.3.3. Second melting events, F2 strongest effect with a tensile strength of 16 MPa was observed for the
DSC scans of the second melting for PBAT/PHB blends and PBAT/ blend 75%PBAT/25%PHB containing the highest PBAT amount. PBAT
PHB/babassu composites are shown in Fig. 10. The endotherm melting is known to behave as a plasticizer making the PHB tougher, i.e. im-
signals show two major overlaid peaks around 150 °C and a minor peak proved energy absorption mechanisms result higher toughness. Blend
around 130 °C for the melting process. The displayed peaks can be as- compositions with 50 and 25% PBAT showed a slightly lower but for
sociated with the neat polymers melting: PBAT melting in the lower both blends equal tensile strength of approx. 10 MPa. Babassu in-
temperature region at roughly 130 °C and PHB melting in the higher corporation up to 3 wt% does not affect Younǵs modulus or elongation
temperature region at 150 °C, which appears as complex peaks [41]. at break of the studied blends. Tensile strength properties show a slight
The temperature range of the endotherms associated to PBATs and tendency to a lower strength upon adding babassu but this effect is low
PHBs second melting increased with the PBAT content. Although ba- (within experimental error), and independent of the babassu content.
bassu addition did not promote changes on PBAT endotherm, a subtle The blend and composites composed of 75%PBAT/25%PHB showed
increase in the temperature range of PHB endotherms was observed an elongation at break of roughly 650%. A comparable effect was ob-
(Fig. 10). Thus, the melting process occurs for the two polymer phases served by França et al. for a PLA/PBAT system [43]. However, the
independently. However, the presence of babassu reduces the melting elongation at break decreases significantly below 3.5% for blends and
temperature to the range of the PBAT phase. composites containing less PBAT. Summing up, blends composed of
Summing up all thermal events in PBAT/PHB blends and 50%PBAT/50%PHB showed the best mechanical properties regarding

6
R. Hoffmann, et al. Composites Part A 124 (2019) 105472

700
16
600

Young's modulus (MPa)

Tensile strength (MPa)

500
12
400

300 8

200 25%PBAT/75%PHB 25%PBAT/75%PHB

4
50%PBAT/50%PHB 50%PBAT/50%PHB
100
75%PBAT/25%PHB 75%PBAT/25%PHB
0 0
0 1 2 3 0 1 2 3
Babassu content (%) Babassu content (%)

1000

25%PBAT/75%PHB
Maximum strain (%)

50%PBAT/50%PHB
100
75%PBAT/25%PHB

10

1
0 1 2 3
Babassu content (%)

Fig. 11. Tensile properties of PBAT/PHB/babassu compounds. Young’s modulus (up left), Tensile strength (up right) and Elongation at break (down). (For inter-
pretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

stiffness and strength independently of babassu's presence and con- torque rheometry, changes in the polymer structure were not significant
centration. enough to alter either the crystallinity or thermal stability of these
systems and, consequently, their tensile properties.
4. Conclusion An overview of the results obtained indicate that the 50%PBAT/
50%PHB composites with 1 wt% and 3 wt% of babassu could be the
On the pursuit of biodegradable alternatives for conventional most suitable composition for packaging and mulch films applications.
packaging polymers PBAT/PHB/babassu compounds with varying They feature lower degradation rates during processing while keeping
composition were rheomixed in an internal mixer and films were ob- crystallinity and promising mechanical performance. This contribution
tained by flat-die extrusion successfully. During processing, degrada- elucidated babassu as an ecologically friendly filler facilitating PHB
tion of the compounds could be observed due to thermal decomposi- processing while maintaining the overall mechanical properties and
tion. Herein, the role of the babassu filler was neglectable whereas the giving access to cost-effective composite materials.
composition of the blend majorly determines the degradation kinetics.
The thermal stability at elevated temperature drops with 1 wt% babassu
and increases slightly when 3 wt% were added. Thus, the overall sta- Acknowledgments
bility during processing is determined by the blend composition.
DSC studies revealed a significant decrease of crystallinity for PBAT The authors would like to thank Conselho Nacional de
rich compounds. Furthermore, the addition of babassu reduces the Desenvolvimento Científico e Tecnológico (CNPq) for the fellowship of
crystallization rate and allows by this for prolonged handling time of Carvalho and the financial support from Deutscher Akademischer
soft films after extrusion prior to solidification. Thus, blend composition Austauschdienst (DAAD) and Coordenação de Aperfeiçoamento de
and babassu addition beneficially affect film-die extrusion processes as Pessoal de Nível Superior (CAPES) within the PROBRAL project PPP
films based on PBAT/PHB and babassu were easier to handle and stayed Brasilien (Projektbezogener Personenaustausch Brasilien). R. Hoffmann
longer flexible in a film roll procedure. gratefully acknowledges funding by the Deutsche
Tensile testing showed that blends composed of 50%PBAT/50%PHB Forschungsgemeinschaft (DFG) under grant number HA 2420/16-1. K.
exhibits the best mechanical properties regarding stiffness and strength, Koschek acknowledges the Bundesministerium für Bildung und
whereas the presence and concentration of babassu does not influence Forschung (BMBF) for funding and support through the NanoMatFutur
significantly the mechanical performance of the blends. award (DuroCycleFVK 03XP0001). João Emídio Neto and Felipe B.
Although some degradation was evidenced during compounding by Calvacanti deserve credit for support in extrusion and mixing.

7
R. Hoffmann, et al. Composites Part A 124 (2019) 105472

Appendix A

A.1. Theoretical background

A.1.1. Rheological characterization

In tests performed in an internal mixer at constant rotor speed, torque is directly proportional to the rate of mechanical energy dissipation. During
the last processing stage (melt processing), and considering the prevailing shear flow inside the mixing chamber, mechanical energy dissipation is
directly proportional to the viscosity η of the molten material [15]. Thus, at any given time:
Z (t ) = kη (t ) (1)
where k is a constant. Viscosity depends on temperature, which varies in time; the temperature dependence of the viscosity may be expressed, in first
approximation by a simple exponential:
η = η ∗exp{ −β (T − T ∗)} (2)
−1
where β (°C ) is the so called temperate coefficient of the viscosity and T is an (arbitrary) reference temperature. The coefficient β should be
*

experimentally determined for the system and processing conditions in question, from previous works [44,45] indicate that a value β ≈ 0.02 °C−1
may be adequate for PHB and PBAT.
Combining Eqs. (1) and (2):
Z = Z ∗exp{ −β (T − T ∗)} (3)
*
Z is a temperature-adjusted torque, a convenient measure of the viscosity at the reference temperature, which may be estimated from the
experimental data Z(t) and T(t) once the coefficient β is known, and used for comparative purposes.
The adjusted torque depends on the composition of the sample, in particular on the molar mass of the polymeric matrix. Almost all known
polymer melts show a remarkably uniform dependence of the viscosity on the mass-average molar mass Mw [14,46], making changes in melt
viscosity – hence, in torque – a very sensitive measure of changes in molar mass.
η ∝ Mw3.5 (4)
Drop in adjusted torque during last stage of melt processing may be associated to incipient polymer degradation during processing in the
laboratory internal mixer. Thus, the relative rate of decrease of the temperature-adjusted torque:
1 dZ ∗
RZ =
¯
Z ∗ dt (5)
is a fair measure of the rate of polymer degradation during processing. When time is measured in minute, 100RZ is just the % drop in torque per
minute of processing time.

A.1.2. DSC characterization

Phase transition events (melting and crystallization) are identified in the DSC scan, their initial and final points established visually, along with
suitable virtual baselines during the events, and the corresponding peaks integrated point by point using custom software. Conversion (fraction
molten or crystallized) x = x(t) was estimated as:
1 t
x (t ) =
E0
∫t 1
|J (t ′) − J0 (t ′)| dt ′
(6)
where J is the heat flow during the phase transition event, J0 is the virtual baseline during the event, and
t2
E0 = ∫t 1
|J (t ) − J0 (t )| dt
(7)
is the total heat transferred between the sample and the neighborhood. The rate of melting/crystallization c = c(t) is:
dx |J (t ) − J0 (t )|
c (t ) ≡ =
dt E0 (8)
Several parameters of the phase transition processes were estimated. In particular, the specific latent heat of melting/crystallization of the blend
was computed as:
E0
ΔH =
wP mS (9)
where mS is the mass of the sample and wP is the mass fraction of the polymers; crystallinity expressed during the event was then estimated as:
ΔH
ΔX =
ΔHm0 (10)
where ΔHm° is the latent heat of melting of the 100% crystalline polymers. For PBAT and PHB 114 J/g [47] and 146 J/g [48], respectively were
applied.
In the DSC scans where PBAT and PHB crystallization/melting peaks where observed as distinct events the degree of crystallinity was measured
for the neat polymers using Eq. (10), in cases where only one peak was observed it was assumed to be the PBAT/PHB blend, and the degree of
crystallinity measured taking account the equilibrium melting temperature according Eq. (11):
(ΔHmo )Blend = wsPBAT (ΔHmo )PBAT+wsPHB (ΔHmo )PHB (11)
where: ws corresponds the fraction of the polymer constituent in the blend or composite. Measured results for (ΔHmo )Blend using Eq. (11) are in Table 3.

8
R. Hoffmann, et al. Composites Part A 124 (2019) 105472

Table 3
Equilibrium latent heat of melting of PBAT/PHB blends.
Blend ΔHm° (J/g)

25%PBAT/75%PHB 138
50%PBAT/50%PHB 130
75%PBAT/25%PHB 122

Appendix B. Supplementary material

Supplementary data to this article can be found online at https://doi.org/10.1016/j.compositesa.2019.105472.

References babassu mesocarp incorporation on the biodegradation of a PBAT/TPS Blend.

Macromol Symp 2019;383(1):1800043.
[25] Arrieta MP, Lopez J, Rayon E, Jimenez A. Disintegrability under composting con-
[1] European Association of Plastics Recycling and Recovery Organisations (EPRO). ditions of plasticized PLA–PHB blends. Polym Degrad Stab 2014;107(16):139–49.
Plastics - The Facts 2017: An Analysis of European Plastics Production, Demand and [26] Beber VC, de Barros S, Banea MD, Brede M, de Carvalho LH, Hoffmann R, et al.
Waste Data. Brussels: Plastics Europe; 2018. Effect of babassu natural filler on PBAT/PHB biodegradable blends. Materials
[2] Jambeck JR, Geyer R, Wilcox C, Siegler TR, Perryman M, Andrady A, et al. Plastic 2018;11:820–35.
waste inputs from land into the ocean. Science 2015;347:768–71. [27] Wojtczak M, Galeski A, Pracella M. Inhibited crystallization of polyhydroxybutyrate
[3] Zumstein MT, Schintlmeister A, Nelson TF, Baumgartner R, Woebken D, Wagner M, by blending with aliphatic-aromatic copolyester. Eur Polym J 2018;103:133–44.
et al. Biodegradation of synthetic polymers in soils: tracking carbon into CO2 and [28] Modi S, Koelling K, Vodovotz Y. Assessment of PHB with varying hydroxyvalerate
microbial biomass. Sci Adv 2018;4(7):eaas9024. content. Eur Polym J 2011;47:179–86.
[4] Laycock B, Nikolic M, Colwell JM, Gauthier E, Halley P, Bottle S, et al. Lifetime [29] Islam MU, Ullah MW, Khan S, Shah N, Park JK. Strategies for cost-effective and
prediction of biodegradable polymers. Prog Polym Sci 2017;71:144–89. enhanced production of bacterial cellulose. Int J Biol Macromol 2017;102:1166–73.
[5] Chiellini E, Solaro R. Biodegradable polymeric materials. Adv Mater [30] Herrera R, Franco L, Rodríguez-Galán A, Puiggalí J. Characterization and de-
1996;8:305–13. gradation behavior of poly(butylene adipate-co-terephthalate)s. J Poly Sci A: Polym
[6] Bastioli C. Handbook of biodegradable polymers. Shrewsbury: Rapra Technology; Chem 2002;40:4141–57.
2005. [31] Dou Q, Cai J. Investigation on polylactide (PLA)/poly(butylene adipate-co-ter-
[7] Shahlari M, Lee S. Mechanical and morphological properties of poly(butylene adi- ephthalate) (PBAT)/bark flour of plane tree (PF) eco-composites. Materials
pate-co-terephthalate) and poly(lactic acid) blended with organically modified si- 2016;9(393):32–45.
licate layers. Polym Eng Sci 2012;52(7):1420–8. [32] Fukushima K, Wu MH, Bocchini S, Rasyida A, Yang MC. PBAT based nanocompo-
[8] Ferreira FV, Cividanes LS, Gouveia RF, Lona LMF. An overview on properties and sites for medical and industrial applications. Mater Sci Eng C 2012;32:1331–51.
applications of poly(butylene adipate-co-terephthalate)–PBAT based composites. [33] Vitorino MBC, Cipriano PB, Wellen RMR, Canedo EL, de Carvalho LH.
Polym Eng Sci 2019;59(S2):E7–15. Nonisothermal melt crystallization of PHB/babassu compounds Kinetics of crys-
[9] Yamamoto M, Witt U, Skupin G, Beimborn D, Müller RJ. Biodegradable aliphatic- tallization. J Therm Anal Calorim 2016;126:755–69.
aromatic polyesters: “Ecoflex®”. Weinheim: Wiley-VCH; 2005. Biopolymers online. [34] Costa ARM, Reul LTA, Sousa FM, Ito EN, de Carvalho LH, Canedo EL. Degradation
[10] Wang H, Wei D, Zheng A, Xiao H. Soil burial biodegradation of antimicrobial during processing of vegetable fiber compounds based on PBAT/PHB blends. Polym
biodegradable PBAT films. Polym Deg Stab 2015;116:14–22. Test 2018;69:266–75.
[11] Pogrell H. Microbes fully metabolize biodegradable plastics. European Bioplastics; [35] Montoro SR, Shigue CY, de Sordi MLT, dos Santos AM, Ré MI. Estudo cinético da
2018. https://www.european-bioplastics.org/microbes-fully-metabolize- redução da massa molar do poli(3-hidroxibutirato-co-3-hidroxivalerato) (PHBHV).
biodegradable-plastics/ [accessed 18 Nov 2018]. Polímeros 2010;20(1):19–24.
[12] Yu L, Dean K, Lin L. Polymer blends and composites from renewable resources. Prog [36] Almeida TG, Neto JES, Costa ARM, Costa AS, de Carvalho LH, Canedo EL.
Polym Sci 2006;31:576–602. Degradation during processing in poly(butylene adipate-co-terephthalate)/vege-
[13] Avella M, Gaceva GB, Buzarovska A, Errico ME, Gentile G, Grozdanov A. Poly(3- table fiber compounds estimated by torque rheometry. Polym Test 2016;55:204–11.
hydroxybutyrate-co-3-hydroxyvalerate)-based biocomposites reinforced with kenaf [37] Mohanty AK, Misra M, Drzal LT. Natural fibers, biopolymers and biocomposites.
fibers. J Appl Polym Sci 2007;104(3):192–200. Boa Raton: Taylor and Francis; 2005.
[14] Dealy JM, Larson EG. Structure and rheology of molten polymers. From structure to [38] Ries A, Canedo EL, Souto CR, Wellen RMR. Non-isothermal cold crystallization
flow behavior and back again. Munich: Hanser; 2006. p. 131–91. kinetics of poly(3-hydoxybutyrate) filled with zinc oxide. Thermochim Acta
[15] Wetzel MD, Shih CK. Experimental simulation with a simple mixer and real mate- 2016;637:74–81.
rial. In: Manas-Zloczower I, editor. Mixing and compounding of polymers. Munich: [39] Feng S, Wu D, Liu H, Chen C, Liua J, Yao Z, et al. Crystallization and creep of the
Hanser; 2009. p. 479–92. graphite nanosheets basedpoly(butylene adipate-co-terephthalate) biocomposites.
[16] Pachekoski WM, Dalmolin C, Agnelli JAM. The influence of the industrial proces- Thermochim Acta 2014;587:72–80.
sing on the degradation of poly(hidroxybutyrate) – PHB. Mater Res [40] Wellen RMR, Rabello MS, Canedo EL. Melting and crystallization of poly(3-hy-
2012;16(2):327–32. droxybutyrate)/carbon black. J Mater Res 2015;30:3211–26.
[17] Arrieta MP, Lopez J, Lopez D, Kenny JM, Peponi L. Effect of chitosan and catechin [41] Wellen RMR, Rabello MS, Fechine GJM, Canedo EL. The melting behaviour of poly
addition on the structural, thermal, mechanical and disintegration properties of (3-hydroxybutyrate) by DSC. Reproducibility study. Polym Test 2013;32:215–20.
plasticized electrospun PLA-PHB biocomposites. Polym Degrad Stabil [42] Costa ARM, Santos RM, Ito EN, de Carvalho LH, Canedo EL. Melt and cold crys-
2016;132:145–56. tallization in a poly(3-hydroxybutyrate) poly(butylene adipate-co-terephthalate)
[18] Przybysz M, Marć M, Klein M, Saeb MR, Formela K. Structural, mechanical and blend. J Therm Anal Calorim 2019. https://doi.org/10.1007/s10973-019-08027-9.
thermal behavior assessments of PCL/PHB blends reactively compatibilized with [43] França DC, Almeida TG, Abels G, Canedo EL, de Carvalho LH, Wellen RMR, et al.
organic peroxides. Polym Test 2018;67:513–21. Tailoring PBAT/PLA/Babassu films for agriculture mulch application. J Nat Fibers
[19] Auriemma M, Piscitelli A, Pasquino R, Cerruti P, Malinconico M, Grizzuti N. 2018;15:1–11.
Blending poly(3-hydroxybutyrate) with tannic acid: influence of a polyphenolic [44] Costa ARM, Almeida TG, Silva SML, de Carvalho LH, Canedo EL. Chain extension in
natural additive on the rheological and thermal behavior. Eur Polym J poly(butylene adipate-co-terephthalate). Inline analysis in a laboratory internal
2015;63:123–31. mixer. Polym Test 2015;42:115–21.
[20] Al-Itry R, Lamnawar K, Maazouz A. Improvement of thermal stability, rheological [45] Marinho VAD, Pereira CAB, Vitorino MBC, Silva AS, de Carvalho LH, Canedo EL.
and mechanical properties of PLA, PBAT and their blends by reactive extrusion with Degradation and recovery in poly(butylene adipate-co-terephthalate)/ thermo-
functionalized epoxy. Polym Degrad Stab 2012;97(10):1898–914. plastic starch blends. Polym Test 2017;58:166–72.
[21] Al-Itry R, Lamnawar K, Maazouz A. Rheological, morphological, and interfacial [46] Doi M. Explanation for the 3.4-power law for viscosity of polymeric liquids on the
properties of compatibilized PLA/PBAT blends. Rheol Acta 2014;53(7):501–17. basis of the tube model. J Polym Sci: Polym Phys 1983;21:667-684.
[22] Al-Itry R, Lamnawar K, Maazouz A. Biopolymer blends based on poly (lactic acid): [47] Gan ZH, Kuwabara K, Yamamoto M, Abe H, Doi Y. Solid-state structures and
shear and elongation rheology/structure/blowing process relationships. Polymers thermal properties of aliphatic-aromatic poly(butylene adipate-co-butylene ter-
2015;7(5):939–62. ephthalate) copolyesters. Polym Degrad Stabil 2004;83:289-300.
[23] Raza ZA, Riaz S, Banat IM. Polyhydroxyalkanoates: properties and chemical mod- [48] Barham PJ, Keller A, Otun EL, Holmes PA. Crystallization and morphology of a
ification approaches for their functionalization. Biotechnol Progr bacterial thermoplastic: poly-3-hydroxybutyrate. J Mater Sci 1984;19:2781–9.
2017;34(1):29–41.
[24] Nunes MABS, Catro-Aguirre E, Auras RA, Bardi MAG, de Carvalho LH. Effect of