JMRTEC-566; No.

of Pages 11
ARTICLE IN PRESS
j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx

Available online at www.sciencedirect.com

www.jmrt.com.br

Original Article

High-density polyethylene/mollusk shell-waste

composites: effects of particle size and coupling
agent on morphology, mechanical and thermal
properties

P.M.A. Melo a , O.B. Macêdo b , G.P. Barbosa b , M.M. Ueki c , L.B. Silva a,b,∗
a Programa de Pós Graduação em Ciência e Engenharia de Materiais, Universidade Federal da Paraíba, Cidade Universitária, S/N,
58051-900, João Pessoa, Paraíba, Brazil
b Departamento de Engenharia de Materiais, Universidade Federal da Paraíba, Cidade Universitária, S/N, 58051 900, João Pessoa,

Paraíba, Brazil
c Departamento de Ciência e Engenharia de Materiais, Universidade Federal de Sergipe, 49100-000, São Cristovão, Sergipe, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: Composites based on high-density polyethylene (HDPE) and mollusk shell-waste (MSW) par-
Received 2 July 2018 ticles, with several concentrations and two different micro-sizes, were prepared by melt
Accepted 8 January 2019 compounding and injection molding. The matrix flow behavior was not altered by the
Available online xxx MSW incorporations. SEM analyses revealed weak filler/matrix interfacing and the pres-
ence of agglomerates (in the composites with higher particle concentrations). For the 2 wt%
Keywords: composites, good filler dispersion and distribution was observed in the HDPE matrix. The
Mollusk shell-waste crystallinity of the matrix increased approximately 10% when adding lower MSW concentra-
Polyethylene tions. For the composites with higher MSW concentrations and finer particles the thermal
Particle size degradation temperature of HDPE (at a 20% mass loss) increased by approximately 26 ◦ C. The
Thermal analysis tensile modulus increased by approximately 10% for the 2 wt% and 8 wt% composites with
Morphology coarse particles. With lower MSW concentrations and finer particles, the flexural modulus
Mechanical properties increased by more than 37%. The composites modified with titanate presented properties
similar to unmodified composites. Molecular interactions were verified by the presence of
a band at 1030 cm−1 , yet were not sufficient to promote improvements in the mechanical
properties of the modified composites.
© 2019 Published by Elsevier Editora Ltda. on behalf of Brazilian Metallurgical, Materials
and Mining Association. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).

∗
Corresponding author.
E-mail: lucineide@ct.ufpb.br (L. Silva).
https://doi.org/10.1016/j.jmrt.2019.01.008
2238-7854/© 2019 Published by Elsevier Editora Ltda. on behalf of Brazilian Metallurgical, Materials and Mining Association. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
2 j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx

an example; silicon dioxide is present in Brazilian mollusk

1. Introduction shells [16], yet is lacking in Korean oyster shells [17]. Along
the coasts of Brazil and other parts of the world there is great
The majority of common household, electronic and auto-
abundance of shell-waste, which is an environmental haz-
motive applications of thermoplastics involve mixtures with
ard lacking storage/disposal. On the Brazilian coast, human
mineral fillers, such as calcium carbonate (CaCO3 ), talc, mica,
consumption drives commercial extraction of mollusks and
wollastonite, glass beads, silica, rice husks and clays. Adding
the eventual production of shell-wastes. Once the mollusk is
such fillers to polymers is very economical and improves the
consumed, the majority of shells are discarded on river banks.
properties of polymeric matrices. Apart from reducing the
The aim of this work was to prepare HDPE/shell-waste
price of the final material, mineral fillers can also help improve
composites using two different grades of mollusk shell par-
stiffness, and reduce both flammability and molding shrink-
ticles and a titanate coupling agent. We investigated tensile,
age, which are principal limitations of bulk thermoplastics
flexural, and impact properties for the composites, together
[1–4]. In polyolefins, calcium carbonate is commonly used as
with flow behaviors, thermal properties, and morphological
filler with particle sizes varying from macro to nano scales.
characteristics to find a new low cost way to recycle Brazilian
Calcium carbonate confers toughness and rigidity to the com-
shell-wastes.
posites [5,6]. HDPE with particulate fillers is of great interest
in industrial applications because it is easily processed, and
it accepts different types of natural and synthetic fillers [7]. 2. Materials and methods
The effects of inorganic filler additions on the mechanical and
thermal properties of polymeric matrices have been widely 2.1. Materials
studied [8–12].
Using a coupling agent to improve the interfacial inter- The polymer used in this study was high-density polyethy-
actions between inorganic fillers and a polymeric matrix, lene (HDPE) IA58 supplied by Braskem. It has a density of
Monte [13] proposed a mechanism to explain interlink- 0.957 g/cm3 and a melt flow index of 22.0 g/10 min (2.16 kg
ing between the inorganic filler, the polymer, and the at 190 ◦ C). Mollusk shell-waste was collected after disposal
titanium derived coupling agents. Briefly, the hydrolysable on the banks of the Paraíba River by a riverside community
portion of the titanate molecule reacts with free protons in Paraiba State, Northeast Brazil. It was used as bio-filler
or hydroxyl groups on the inorganic surface, leading to for this study. The organic titanate coupling agent (isopropyl
formation of organic monomolecular layers. As a result, dis- tri(dioctylpyrophosphate)) is an industry grade powdered
persion of inorganic filler in an organic phase is enhanced product, and was kindly donated by the Mag-Mix Company.
due to compatibility at the interface: substituting hydrating This specific coupling agent was chosen due to its potential
water at the inorganic surface with a monomolecular layer to improve the thermal stability, and mechanical-rheological
before establishing titanate coupling. Consequently, a com- properties of polyolefinic matrices.
patible inorganic/polymer interface promotes elimination of
unwanted air voids in the composite, and dispersion of filler 2.2. Mollusk shell preparation
into the matrix.
Researchers around the world [14,15] have studied the The shell-wastes were washed to remove superficial impuri-
structural features of different types of mollusk shells in ties and then dried outdoors. They were then milled on a disc
great detail. Most shells are natural composites composed mill (30 min) and sieved. Granulometric classification was per-
of stacked platelets (0.5 ␮m) arranged as if they were a brick formed using two different sieves: 200 (74 ␮m) and 325 (44 ␮m)
and mortar microstructure, with an organic matrix inter- mesh; the shell powder was respectively classified into two
layer (20–50 nm). The crystalline structure of mollusk shell grades: coarse and fine.
is composed of aragonite and calcite, both are polymorphs
of CaCO3 , and its chemical composition is 97.5372% calcium
2.3. Processing
oxide [16]. Similar findings have been reported for other types
of seashells [15,17].
After the shell-waste was washed, dried and sieved, one part
Recently, the need to develop materials that are easily recy-
of the shell-waste was mixed to 1.2% (m/m) with powdered
clable and that contribute less to greenhouse gas emissions
titanate coupling agent. Using the coarse and fine grade com-
has motivated research involving wood residues and shells
posites, we then prepared 2, 5, and 8% (m/m) shell-waste
as bio-fillers for polymeric matrices [17,18]. This motivated
particle mixtures. Composites using 5% (m/m) titanate modi-
us to investigate HDPE polymeric bio-composites with recy-
fied shell-waste were also studied. The samples were melted
cled mollusk shell-waste particles. The use of treated shells
into a compound using a 30 mm fluted single screw extruder
for polyolefins with cetyltrimethylammonium bromide [17],
(as proposed by Tadmor and Gogos [21]). The extrusion mix-
furfural [19], furfural and hydrochloric acid [20] has been well
ing process was performed at a temperature of 170–180 ◦ C,
documented, as tests of polymeric composites mixed using
with screw rotation speed at 50 rpm. A fluted screw was pre-
twin-screw extruders [17,19,20]. Varied authors have studied
ferred because a single screw extruder is less expensive than
seashells as bio-fillers in polymeric composites. The peculiar
a twin screw extruder, making processing more economic
properties of these materials depend largely on processing,
and attractive for recycling companies. The preparation of
shell-waste surface treatments, and the polymeric matrix. In
the test specimens from the compound composites occurred
addition, shells differ in respect to chemical composition. As
in a BATTENFELD HM45/210 injection molding machine. The

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx 3

constant integration limits. The degree of crystallinity (Xc)

Table 1 – Sample labeling.
was determined using the following equation:
Designation Samples

MSW Mollusk shell-waste particles Hm

Xc (%) = 100 (1)
MSH2 2.0% MSW (200 mesh) (1 − Wf ) · H100%
MSL2 2.0% MSW (325 mesh)
MSH5 5.0% MSW (200 mesh)
MSL5 5.0% MSW (325 mesh) where Hm is the melting enthalpy per unit weight of HDPE
MSH5T 5.0% MSWT (200 mesh) in the composition; Wf is the fractional weight of shell-waste
MSL5T 5.0% MSWT (325 mesh) particles in the composite; and H100% denotes the enthalpy
MSH8 8.0% MSW (200 mesh) per unit weight of 100% crystalline HDPE, which was assumed
MSL8 8.0% MSW (325 mesh) to be 293 J/g [22].

2.4.5. Thermogravimetric analysis (TGA)

molding conditions were as follows: injection speed 75 cm/s;
The curves were obtained using a Shimadzu DTG-60H
holding pressure 600 bar; hold time 15 s; and cooling time 30 s.
Simultaneous DTA-TG apparatus. The sample weights were
The temperature profile used was 170–190 ◦ C, and the mold
approximately 10–11 mg, and samples were loaded and heated
temperature was maintained at 30 ◦ C. Before processing, the
from 30 to 900 ◦ C at 10 ◦ C/min in an argon atmosphere.
materials were dried in an air circulating oven for 2 h at 60 ◦ C.
The particulate compositions are shown in Table 1.
2.4.6. Scanning electron microscopy (SEM)
The morphologies of the composites were investigated using a
2.4. Characterization
ZEISS LEO 1430 scanning electron microscope; observing dis-
persion and interfacial adhesion of the mollusk shell-waste
2.4.1. Particle size distribution
particles to the HDPE matrix. The injected composite speci-
The particle size distribution curves for both particle samples
mens were immersed in liquid nitrogen, after several minutes
were obtained by laser diffraction using a Cilas laser diffrac-
fractured cryogenically, and then covered with a thin layer
tion analyzer in wet mode with distilled water.
of gold for analysis of the fracture surface. The images were
obtained with different magnifications at an electron beam
2.4.2. Fourier transform infrared spectroscopy (FTIR) accelerating voltage of 15 kV.
The analyses were recorded on an IR-Prestige 21 spectropho-
tometer from Shimadzu, accumulating 32 scans with 4 cm−1
resolution in the 4000 to 1000 cm−1 region. The samples
2.4.7. Mechanical properties
Tensile, flexural, and impact tests were performed on the
of mollusk shell-waste and titanate coupling agent using
HDPE/mollusk shell composites to analyze the effects on the
approximately 1 wt% were prepared as KBr pellets. Analyses
mechanical performance of adding the differently sized mol-
of the ground composite samples were performed in attenu-
lusk shell-waste particles. For each compound, five samples
ated total reflectance (ATR) mode by direct analysis on a ZnSe
were tested, and mean values were considered. The tensile
crystal. The FTIR spectra were baseline and ATR-corrected.
tests were carried out using a universal testing machine; Shi-
madzu model AG-X 10KN, according to ASTM D 638 at an
2.4.3. Melt flow
extension rate of 50 mm/min. The three-point bend flexural
The processability of the filled HDPE composites was deter-
test was conducted conforming to ASTM 790 using the same
mined with a Modular Line CEAST melt flow indexer according
universal testing machine; Shimadzu model AG-X 10KN, at
to ASTM 1238-10. A load of 2.16 kg at 190 ◦ C was used for the
a deflection rate of 2 mm/min. Izod impact tests were per-
measurements. At least five specimens for each composition
formed in accordance with ASTM D256 using a CEAST Resil
were tested. Before analysis, the samples were dried in an air
5.5 machine and a 2.75 J pendulum impact tester.
circulating oven for 2 h at 60 ◦ C.

2.4.4. Differential scanning calorimetry (DSC) 3. Results and discussion

Thermograms were obtained using a Shimadzu DSC 60
calorimeter. Calibration was done with indium and sapphire
in the temperature range of 0–350 ◦ C. The sample weight was
approximately 6.0–6.4 mg. For each test, the sample was first
heated to 200 ◦ C at 10 ◦ C/min and then annealed for 5 min to
ensure an identical thermal history. The sample was subse-
quently cooled to 30 ◦ C at a cooling rate of 10 ◦ C/min for data
collection. In the second heating scan, the sample was heated
from 25 to 200 ◦ C at 10 ◦ C/min in a nitrogen atmosphere.
The crystallization temperature (Tc) was determined during
cooling, and the peak temperature of the second heating scan
was taken as the melting temperature of the polymer. The
peak area was used to determine the melting enthalpy using
In this work, composites with low concentrations of shell-
waste were investigated. The single fluted screw extruder in
laboratory scale did not permit mixing with a shell-waste
concentration above 8% (m/m). Chong et al. [17] has studied
recycled polyethylene/shell-waste composites with various
higher concentrations without observing changes relevant
to the mechanical matrix properties. Though other authors
[19,20] have employed pretreatment for shell-wastes using a
sodium hydroxide solution, this was not done in the present
study. Our proposal was to prepare low cost composites
without shell-waste pretreatments, and to investigate the
influence of low concentration shell-wastes (in two different

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
4 j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx

100 2.4 4000 3000 2000 1000 0

325 mesh
100
2.1

80 80
1.8
Cumulative value (%)

Histogram (%)
1.5 60
60 Titanate agent
D50 = 5.27µm
1.2
100
40 0.9
95
90
0.6 85
20

Transmittance
80
MSL5T
0.3 75
100
0 0.0
1 10 100 1000 90
Diameter (µm)
80
MSL5
Fig. 1 – Particle size distribution of mollusk shell-waste 70
(325 mesh samples). 100

95

90
particle sizes) on the flow, thermal, mechanical, and morpho- MSH5T
logical characteristics of polyethylene/bio-filer compositions. 85
100

3.1. Particle size distribution 95

90 MSH5
The particle size distribution curve for the 325 mesh sample
is shown in Fig. 1; the results for the 200 mesh sample had 4000 3000 2000 1000 0
already been investigated in a previous study [16]. Comparing
Wavenumber (cm-1)
the results, the 325 mesh sample presented a 50% diameter
(D50) of 5.27 ␮m, this was compared to the 50% diameter (D50) Fig. 2 – FTIR spectra for the titanate coupling agent, MSH5,
of 6.49 ␮m obtained from the 200 mesh sample, resulting in a MSH5T, MSL5 and MSL5T composites in the 4000 to
diameter decrease of approximately 23%. The same tendency 1000 cm−1 region.
was observed for particles with diameters at 90% (D90); result-
ing in a diameter decrease of approximately 28%.
Table 2 – MFI for pure HDPE and its composites.
3.2. FTIR
Designation MFI (g/10 min)

Fig. 2 presents the FTIR spectra of the titanate coupling agent HDPE 22.0 ± 0.50
and composites, detailing the interaction between mollusk MSH2 21.8 ± 4.01
MSL2 21.7 ± 2.89
shell-waste and HDPE in composites with 5 wt% concen-
MSH5 22.5 ± 1.28
trations. The mollusk shell exhibits absorptions bands at
MSL5 22.7 ± 0.98
1475 cm−1 , 1080 cm−1 , and 858 cm−1 and a split peak at MSH5T 22.4 ± 2.08
710 and 700 cm−1 being characteristic bands of aragonite MSL5T 22.7 ± 1.56
[23]. The titanate coupling agent spectra present absorption MSH8 20.8 ± 2.65
bands at 2930 cm−1 and 2850 cm−1 , which are respectively MSL8 21.8 ± 3.98
related to CH2 and CH3 group stretching. Additionally, a broad
absorption band at 3500–3100 cm−1 was observed, which is 3.3. Melt flow index (MFI)
characteristic of OH groups. The bands at 1720 cm−1 and
1462 cm−1 are attributed, respectively, to C O stretching vibra- Table 2 shows the melt flow rates (2.16 kg at 190 ◦ C) of the
tion and bending vibrations of methylene groups. The sharp composites previously presented in Table 1. Adding MSW to
and strong band at 1030 cm−1 is related to the Ti-O-C linkage the HDPE (up to 8 wt%) did not significantly alter the melt-
of the titanate coupling agent. The composite spectra present ing flow rate. Thus, the mollusk shell particle size variations
the characteristic bands of polyethylene related to the methy- (on the micron scale) with diameters at 50% (between 5.27 ␮m
lene groups near 2900 cm−1 and 1470 cm−1 and a strong sharp and 6.49 ␮m) did not significantly modify the flow rates of
doublet in the 730–700 cm−1 region. Looking at the region of the composites. This behavior is completely different from
1300–800 cm−1 , the composite spectra present a band near other observed nano-particulate composites. For example,
858 cm−1 related to the mollusk shell particles. The band at certain authors [26,27] have observed a considerable increase
1030 cm−1 was evident only in the spectra of composites mod- in the melting flow rate for thermoplastic/nano-CaCO3 , which
ified with the titanate coupling agent, and may also be related was attributed to nano-particle rolling (bearing-like), which
to the Ti–O–Ca linkages, as has been suggested for ZnO [24] reduced interlayer interactions in the nanocomposites when
and CoZ ferrite particles [25]. melted. Jang, Keon-Soo [28] observed that large talc particles

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx 5

a Table 3 – DSC test results for HDPE and its composites.

60
Sample Tconset (◦ C) Cooling 2nd heating
50
◦ ◦
Tc peak ( C) Tm ( C) DHm (J/g) Xc (%)
40 HDPE

MSH2 HDPE 106.70 114.79 131.69 150.6 44.18

Heat Flow

30 MSH2 110.59 114.80 131.98 155.8 48.80

MSH5 MSL2 107.52 114.71 132.05 141.9 46.33
20
MSH5 108.59 114.20 130.65 126.80 46.80
10 MSH5T MSL5 109.62 114.62 131.05 125.90 45.33
MSH8
MSH5T 107.34 114.36 133.14 118.20 42.56
0
Exo MSL5T 109.84 113.95 133.78 133.40 48.02
-10 MSH8 106.44 114.24 132.14 116.2 45.10
0 25 50 75 100 125 100 175 200 225 MSL8 106.39 114.44 131.98 139.4 43.02
Temperature (ºC)
b
30 shell-waste particle concentration or all of the compositions
studied. None of the particle sizes used for filler partici-
HDPE
20 pated in HDPE nucleation, even when the initial crystallization
MSH2
temperature (T(onset) ) was increased for the composite con-
10 MSH5
Heat Flow

centrations below 5 wt%. The increase in T(onset) did however

MSH5T
0
MSH8
-10
Endo
-20
0 25 50 75 100 125 150 175 200
Temperature (ºC)
Fig. 3 – DSC curves for pure HDPE and composites using
two differing particle sizes. Crystallization temperature: (a)
MSL, (b) MSH and melt temperature in second heating: (c)
MSL, (d) MSH.
act as flaws, which due to their flattened shape tend to slide,
yet they reduce the melt flow in composites with higher con-
centrations. The mollusk particles do not present a definitive
shape (as shown in the Figs. 5 and 6). Further, the compos-
ites were studied without addition of any flow aids, and the
maximum MSW concentration used in the composites was
limited to 8 wt%. However, recycling of these wastes as filler
for polymeric composites is possible.
3.4. Differential scanning calorimetry (DSC)
To study the effect of particle size on the thermal proper-
ties of HDPE composites, DSC studies of the non-isothermal
crystallization and melting behavior were performed. A DSC
cooling scan of the exothermic transition was performed after
heating and annealing for 5 min. Fig. 3a shows the DSC cool-
ing scan of the exothermic transition; Fig. 3b shows the DSC
heating scan obtained after cooling under controlled condi-
tions. The thermograms of both composites with small (44 ␮m)
and large (74 ␮m) particles were similar, thus, only those
with large particles are showed in the Fig. 3. Calculating the
degree of crystallinity was achieved using Eq. (1). The melting
enthalpy was obtained in the second heating (Fig. 3b). Table 3
shows the values for crystallization and melting tempera-
ture, melt enthalpy, and degree of crystallinity for all of the
samples. The HDPE crystallization temperature (Tc) was not
altered, remaining at approximately 114 ◦ C, for every mollusk
accelerate the crystallization process for these composites.
The MSW additions did not alter the melt behavior of HDPE,
and the influence of particle size was insignificant. However,
the degree of crystallinity for the MSH2 and MSL5T compos-
ites increased by approximately 10% with respect to that of
225 the pure polymer. This increase was not supported by the
Tc(peak) results, as the values did not change. Nevertheless,
crystallization was initiated in the MSH2 composite at a tem-
perature above 4 ◦ C (with respect to that of the pure polymer),
which could have led to modifications in its crystal struc-
ture. Consequently, these potential changes were insufficient
to increase either Tc or Tm, but may have contributed to crys-
tallinity increases in the composites. Bartczak et al. [7] have
demonstrated that it is possible to obtain higher crystallinity
for HDPE/calcium carbonate (CaCO3 ) composites using a slight
increase in Tc. This is attributed to a reduction in the thick-
ness of the HDPE lamellae, and formation of a new fraction
of very thin crystallites within it. Further, the effects of inor-
ganic fillers as nucleation agents in thermoplastic matrices
have been well documented in the literature for other compos-
ites filled with nano or micro-particles (among them, calcium
carbonate) [7,11,12,26,29]. Comparing the effects of particle
size and coupling agents on T(onset) and degree of crystallinity
for the MSL5T and MSH5T composites, it was observed that
the composites, when using the 44 ␮m particle size filler, pre-
sented better results than with 74 ␮m particles. In fact, the
results reveal that by increasing the specific MSW area, (by
decreasing the particle size to 44 ␮m), the coupling agent was
more efficient. When the concentration of the mollusk shell
particles increased (in the MSH8 and MSL8 composites), no
significant changes were observed in their thermal character-
istics in relation to HDPE.
3.5. Thermogravimetric analysis (TGA)
To investigate the effect of shell-waste filler particle size on
the thermal stability of the HDPE matrix, TGA measurements
were performed. Both components showed only one stage of
decomposition. The temperatures obtained at mass losses of
20% and 50% are shown in Table 4. At 20% mass loss, the
respective pure polymer and MSW temperatures were equal

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
6 j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx

MSL5T temperature was practically unchanged in comparison

Table 4 – TGA test results for HDPE and its composites.
to the MSL5 composite. The maximum degradation tempera-
Sample Temperature at 20% Temperature at 50%
tures of these composites were not altered. The composites
mass loss (◦ C) mass loss (◦ C)
with an 8 wt% MSW concentration, did not present the same
HDPE 434.13 469.98 tendency toward increased thermal stability of the matrix as
MSH2 456.46 469.15 did the other compositions (2 wt% and 5 wt%). From morpho-
MSL2 458.11 471.27
logical observations (Fig. 7), the MSH8 and MSL8 composites
MSH5 462.57 469.15
contained agglomerated particles, though the lower MSW con-
MSL5 455.05 472.68
MSH5T 448.23 469.81 centration composites presented better particle distributions
MSL5T 460.92 471.27 (Fig. 5). The improvement in thermal stability observed for
MSH8 412.75 469.15 HDPE upon addition of MSW is similar to that of other polymer
MSL8 460.12 471.27 composites obtained with other types of mineral particles,
MSW 711.19 715.00 suggesting that thermal stability is enhanced when filler par-
ticles are well dispersed into the polymer matrix, impeding
a diffusion of volatile decomposition products due to the for-
100 mation of carbonaceous char. Consequently, the filler has a
physical barrier effect, hindering heat diffusion and degrada-
80 tion of volatiles from the bulk [7,11,32–34]. The agglomeration
of MSW in the MSH8 and MSL8 composites may have blocked
such particles, preventing their action as physical barrier to
TG (%)

60 Mollusc shell
HDPE the HDPE decomposition.
MSH2
40
MSH5
MSH5T 3.6. Morphological aspects (SEM)
20 MSH8

Figs. 5–8 present fracture surfaces and cross-sectional areas

0 of the composites. The shell-waste particles are irregular in
0 100 200 300 400 500 600 700 800 900
shape; and in the composites with lower filler concentrations,
especially the 2% composites are well distributed and dis-
Temperature (ºC)
b persed as shown in Fig. 5. Another important observation is
0.15 the presence of voids in the interfacial region of all unmodified
HDPE composites (Figs. 5 and 6). In addition to the voids observed
0.10
MSH2 around the particles, parts of their surfaces were attached
DTG(%/min)

0.05
0.00
-0.05
0 100 200 300 400 500
Temperature (ºC)
Fig. 4 – TGA curves for pure HDPE and MSH composites
with concentrations of filler at 2, 5, and 8 wt%.
to 434.13 ◦ C and 711.19 ◦ C, the 50% mass loss degradation
temperatures were respectively 469.98 ◦ C and 715.00 ◦ C. The
composites presented thermal behaviors similar to that of
HDPE for single-stage decomposition. The thermal degrada-
tion profile of the mollusk shell-waste was similar to that of
CaCO3 , as evaluated by other authors [30,31]; in which the
CaCO3 mass loss stage was attributed to its decomposition
into calcium oxide and carbon dioxide. It can be seen in Table 4
that addition of MSW to HDPE (in both particle sizes) increased
thermal stability. This is attributable to MSW’s higher thermal
stability (being an inorganic filler), as compared to the pure
polymer. Due to the similarity of results for both the 44 ␮m
and 74 ␮m particle sizes, Fig. 4a only presents the curves of
the MSH composites. The temperature at a 20% mass loss
for MSH5 was approximately 14 ◦ C higher than MSH5T. The
MSH5 to the polymer matrix, indicating that the particles were (in
some cases) mechanically anchored. However, regardless of
MSH5T
particle size, when the concentration was increased a strong
MSH8
tendency toward agglomeration could be observed (as shown
by the black arrows) as seen in Fig. 7a and b. In these com-
posites, the particles are only held in place by a thermal
mismatch of the matrix during thermal processing, which
600 700 800 900 1000
induces a clamping effect. The morphological aspect of these
composites corroborates the work of Wang et al. [35] with
hydroxyapatite-polyethylene composites. Under static load-
ing conditions, the weak interaction between the shell-waste
and the HDPE matrix in such composites may not be sufficient
to transfer the stress from the matrix to the particles. From
this perspective, an improvement in the composite tensile
strength would not be expected from higher concentrations
of MSW.
Other authors [36] have observed agglomerates in the films
of polypropylene composites prepared by casting with egg
shell at 5 wt%, being this, the concentration limit for incorpo-
ration of such particles into the matrix resulting in mechanical
property gains. One can observe the effect of the type of pro-
cessing employed, in relation to dispersion and distribution
of the shell particles in the polymer matrices. For example,
egg shell composites prepared by casting process [36], snail
composites prepared using twin-screw extruder [37], and in
the present work, mollusk shell composites using fluted single
screw. From these studies [37], it was concluded that mixtures
are more effective when done using a twin-screw extruder;

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx 7

Fig. 5 – SEM micrographs of the HDPE/MSW fracture surface: (a, b) MSH2 and (c, d) MSL2.

Fig. 6 – SEM micrographs of the HDPE/MSW fracture surface: (a,c) MSH5; (b,d) MSL5.

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
8 j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx

Fig. 7 – SEM micrographs of the HDPE/MSW fracture surface: (a) MSH8; (b) MSL8.

Fig. 8 – SEM micrographs of the HDPE/MSW fracture surface: (a) MSH5T; (b) MSL5T.

even if the composites have higher concentrations of in natura
snail shell.
In this work, isopropyl tri(dioctyl pyrophosphate) titanate
coupling agent was used. It is important to consider that
beyond its main function of forming a monomolecular layer,
and consequently a compatible interfacial region between the
shell-waste and HDPE matrix; its chemistry can also improve
the thermal properties of HDPE/Shell-waste composites. As to
the modified composites (MSH5T and MSL5T) with the titanate
coupling agent, it is suggested that did not occur interfacial
interaction between the modified shell-waste particles and
the HDPE matrix. The presence of 1030 cm−1 band (Fig. 2) did
not favor this interaction. The morphological details can be
observed in Fig. 8a and b. In comparing Figs. 6a, b and 8, it can
be seen that the voids around the particles were smaller, as
seen in Fig. 8 for both modified composites.
3.7. Mechanical properties
The mechanical test results for both unmodified HDPE and its
composites are presented in Figs. 9 and 10. Both of the MSH2
and MSL2 composites presented better mechanical behavior,
as there was an improvement of approximately 7% in yield
tensile values as compared to HDPE, as seen in Fig. 9a. From
a morphological aspect, the good dispersion, and distribution
of the particles together with the mechanical anchorage con-
tributed to impeding fissure propagation, making MSH2 and
MSL2 more resistant. The irregular shapes of the mollusk shell
particles helped their anchorage to the matrix, despite the
voids present around them. Regarding of the presence of parti-
cle agglomerates in the MSL5 and MSL8 composites, the larger
surface areas associated to the fine (44 ␮m) particles con-
tributed to the yield tensile values making them practically the
same as HDPE. However, a decrease by about 7% for this prop-
erty occurred for the MSH5 and MSH8 composites with coarse
(74 ␮m) particles, as they were more affected by the presence
of particle agglomerations. Thus, in accordance with other
authors [36,38], large particle agglomerates affect the ability
of particulate composites to withstand loading. Load trans-
fer from the matrix to the particles occurs more effectively in
composites with higher specific area particles. Figs. 6 and 7
present debonding and pull out of particles, which took place
in the composites (as indicated by red arrows). Accordingly,
composites with finer particles present better results than
those with coarser particles. Further, the coupling agent was
not sufficient to improve the yield tensile of the composites
modified with shell particles at 5 wt%. As shown in Fig. 10, for
tensile modulus, the 2 wt% and 8 wt% composites presented
improvements as compared to the HDPE matrix. It can be seen,
that tensile modulus increases with particle concentration,
the best results were observed with the 8 wt% composites.
It is known from the literature that the tensile modulus of
a polymeric matrix improves with the addition of rigid parti-
cles because the rigidity is generally much higher than that

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx 9

a b
24
600

20
500

Yield tensile (MPA)

Tensile modulus (MPA)

16
400

12
300

8
200

4
100

0
HDPE MSH2 MSL2 MSH5 MSL5 MSH5T MSH5T MSH8 MSL8 0
HDPE MSH2 MSL2 MSH5 MSL5 MSH5T MSH5T MSH8 MSL8

Sample Sample

c d
24 14
22
12
Tensile strength at break (MPA)

20

Elongation at break (%)

18
10
16
14 8
12
10 6

8
4
6
4 2
2
0 0
HDPE MSH2 MSL2 MSH5 MSL5 MSH5T MSH5T MSH8 MSL8
HDPE MSH2 MSL2 MSH5 MSL5 MSH5T MSH5T MSH8 MSL8

Sample Sample

Fig. 9 – Mechanical properties of HDPE and its composites: (a) Yield tensile; (b) tensile modulus; (c) tensile strength at break;
(d) elongation at break.

of the organic polymers [39]. In this study, the composites the composites improved when the particle size was 44 ␮m.
with coarse particles presented higher tensile modulus val- However, a decrease of approximately 23% was observed for
ues than those with finer particles. In Fig. 9c one can observe the MSH5 and MSH8 composites. Thus, coarse particles did
that tensile strength at break for the 2 wt% composites fol- not contribute to the impact properties of these composites.
lowed the same tendency as presented for yield tensile. It The results are in accordance with those of Bharathiraja et al.
is noted that the composites with finer particles presented [38], whose composites (with particle agglomerates) had their
more strength than those with coarser particles (where val- capacity to carry load affected, reducing their impact strength.
ues were reduced in relation to the value of pure HDPE). The The fine mollusk shell particles played an important role in
mollusk shell particle sizes influenced the mechanical prop- improving the flexural and impact mechanical properties of
erties of the HDPE composites. The tensile strength at break the HDPE/MSW composites.
of MSH8 was more affected by aggregated particles and voids
around them, where dewetting of particles occurred more eas-
ily, leading these composites to early failure. It is common in 4. Conclusion
particulate composites with agglomerates that failures begin
with structural inhomogeneity of the material [28]. HDPE com- In this study, mollusk shell-waste particles were added to
posite elongation at break was less affected by mollusk shell HDPE in concentrations of 2, 5 and 8 wt%, using micropar-
particle size and the 8 wt% composites were only negligibly ticles of two different sizes. Superficial modification of the
decreased. particles was performed using a titanate coupling agent.
Under flexural stress (as shown in Fig. 10a and b), both Matrix flow behavior remained unaltered by the mollusk par-
flexural strength and modulus were improved for all compos- ticle additions. Composites with lower concentrations of filler
ites when adding finer particles (44 ␮m). As an example, an presented increased crystallinity when compared to HDPE.
increase in flexural modulus of about 36% was obtained for The degradation temperature (20% mass loss) of the matrix
the MSL2 composite because of excellent particle dispersion increased together with filler concentrations; however, finer
and distribution. For the MSL5 and MSL8 composites, the particles were more effective in this respect. In general,
mechanical properties were only mildly improved in relation the composites with titanate modified particles presented
to HDPE. Despite the morphological aspects of these compos- better thermal properties. The results were collaborated by
ites, (with particle agglomerates), their higher crystallinity (as molecular interactions verified by the presence of a band
compared to pure matrix) may have overcome the negative at 1030 cm−1 , that contribute to decrease void sizes around
effects of particle agglomeration. The impact strength of the particles. With the addition of lower concentrations of

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
10 j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx

24 promotes improvement in both mechanical and thermal prop-

erties. Thus, recycling of this residue as a bio-filler for HDPE is
promising, especially in ecological applications such as gar-
dens. In addition, its use as a bio-filler in other polymeric
Flexura strength (MPa)

16
matrices may help mitigate environmental problems on the
banks of the Paraiba River.

8 Conflicts of interest

The authors declare no conflicts of interest.

0
HDPE MSH2 MSL2 MSH5 MSL5 MSH5T MSL5T MSH8 MSL8
Acknowledgements
Sample
The Authors gratefully acknowledge the financial/technical
support of the following institutions: for the scholarship pro-
1200
vided by the Coordination for Improvement of Higher Level
1000 Education and Personnel (CAPES), and for financial support
Flexura modulus (MPa)

given to project number 476514/2010-0 as provided for by the

800 National Council for Scientific and Technological Develop-
ment (CNPq), and for technical support as provided for by the
600
Fuel and Materials Laboratory (LACOM) and the Fast Solidifi-
cation Laboratory (LSR) at the Federal University of Paraiba.
400

200 references

0
HDPE MSH2 MSL2 MSH5 MSL5 MSH5T MSL5T MSH8 MSL8

Sample [1] Karrad S, Lopes Cuesta JM, Crespy A. Influence of a fine talc
on the properties of composites with high density
40
polyethylene and polyethylene/polystyrene blends. J Mater
Sci 1998;33:453–546.
[2] Ji-Zhao L. Reinforcement and quantitative description of
30 inorganic particulate-filled polymer composites.
Impact strength (j/m)

Composites: Part B 2013;51:224–32.

20
10
0 and its application as fillers for polymers. Mater Sci Eng A
HDPE MSH2 MSL2 MSH5 MSL5 MSH5T MSL5T MSH8 MSL8
Sample
Fig. 10 – Mechanical properties of HDPE and its composites:
(a) flexural strength; (b) flexural modulus; (c) impact
strength.
filler, the yield tensile strength of the matrix increased by
approximately 7%. The tensile modulus increased in the 2 wt%
and 8 wt% particulate composites. Dispersion and distribution
in the lower concentration composites was more effective;
improving the mechanical tensile properties. Yet, compos-
ite elongation at break was practically unchanged, especially
when finer particles were used. Strength and modulus under
flexural strength testing, and impact strength were more
affected by particle size; the composites with finer parti-
cles presented better performance. For HDPE, adding mollusk
shell-waste at concentrations of 2 wt%, (with finer particles)
[3] Jeanette G, Carmem A, Miren I, Berenice D. Effects of
coupling agents on mechanical and morphological behavior
of the PP/HDPE blend with two different CaCO3 . Eur Polym J
2002;38:2465–75.
[4] Lou J, Harinath V. Effects of mineral fillers on polystyrene
melt processing. J Mater Process Technol 2004;152:185–9.
[5] Gorna K, Hund M, Vucak M, Grohn F, Wegner G. Amorphous
calcium carbonate in form of spherical nanosized particles
2008;477:217–25.
[6] Thio YS, Argon AS, Cohen RE, Weinberg M. Toughening of
isotactic polypropylene with CaCO3 particles. Polymer
2002;43:3661–74.
[7] Bartczak Z, Argon AS, Cohen RE, Weinberg M. Toughness
mechanism in semi-crystalline polymer blends: II
High-density polyethylene toughened with calcium
carbonate filler particles. Polymer 1999;40:2347–65.
[8] Cho J, Joshi MS, Sun CT. Effect of inclusion size on
mechanical properties of polymeric composites with micro
and nano particles. Compos Sci Technol 2006;66:1941–52.
[9] Inbakumar JP, Ramesh S. Mechanical, wear and thermal
behaviour of hemp fibre/egg shell particle reinforced epoxy
resin bio composite. Trans Canadian Soc Mech Eng
2018;42:280–5.
[10] Bouza R, Marco C, Naffakh M, Barral L, Ellis G. Effect of
particle size and a processing aid on the crystallization and
melting behavior of iPP/red pine wood flour composites.
Compos Part A: Appl Sci Manufact 2011;42:935–49.
[11] Luyt AS, Dramicanin MD, Antic Z, Djokovic V. Morphology,
mechanical and thermal properties of composites of

Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
JMRTEC-566; No. of Pages 11
ARTICLE IN PRESS
j m a t e r r e s t e c h n o l . 2 0 1 9;x x x(x x):xxx–xxx
polypropylene and nanostructured wollastonite filler.
Polymer Testing 2009;28:348–56.
[12] Alaei MH, Mahajan P, Brieu M, Kondo D, Rizvi SJA, Kumar S,
et al. Effect of particle size on thermomechanical properties
of particulate polymer composite. Iran Polymer J
2013;22:853–63.
[13] Monte SJ, Sugerman G. In: Kresta JE, editor. the use of titanate
coupling agents for improved properties and aging of plastic
composites and coating. Boston: Springer; 1984. p. 301–41.
[14] Meyers MA, Chen PY, Lin AYM, Seki Y. Biological materials:
structure and mechanical properties. Prog Mater Sci
2008;53:1–206.
[15] Pokroy B, Fitch AN, Lee PL, Quintana JP, Caspi EN,
Zolotoyabko E. Anisotropic lattice distortions in the
mollusk-made aragonite: a widespread phenomenon. J
Struct Biol 2006;153:145–50.
[16] Firmino HCT, Chagas TF, Melo PMA, Silva LB.
Characterization of the particulate composites of high
density polyethylene/mollusk shell powder. Matéria
2017;22:1–12.
[17] Mi HC, Byoung CC, Yong-Chan C, Bong GC. Fire-retardant
plastic material from oyster-shell powder and recycled
polyethylene. J Appl Polymer Sci 2006;99:1583–9.
[18] Mosab K, Ko YG. Melt flow behavior and processability of
polylactic acid/polystyrene (PLA/PS) polymer blends. J
Polymer Environ 2017;25:994–8.
[19] Yao ZT, Chen T, Li HY, Xia MS, Ye Y, Zheng H. Mechanical and
thermal properties of polypropylene (PP) composites filled
with modified shell-waste. J Hazard Mater 2013;262:2012–7.
[20] Zhitong Y, Meisheng X, Liugin G, Tao C, Haiyan L, Ying Y,
et al. Mechanical and thermal properties of polypropylene
(PP) composites filled with CaCO3 and shell-waste derived
bio-fillers. Fibers Polymers 2014;15:1278–87.
[21] Tadmor Z, Gogos CG. Principles of polymer processing. 2nd
edn New Jersey: John Wiley &Sons; 2006. p. 235–98 [chapter 6,
Pressurization and pumping].
[22] Krevelen DW. Properties of polymers. New York: Elsevier;
1990.
[23] Zhuona Z, Yidong X, Xurong X, Haihua P, Ruikang T.
Transformation of amorphous calcium carbonate into
aragonite. J Crystal Growth 2012;343:62–7.
[24] Lei Z, Min Z, Hongliang G. Surface modification of zinc oxide
nanorods for potential applications in organic materials.
Appl Surf Sci 2011;258:1551–4.
[25] Hsing-I H, Chih-Cheng C, Jaw-Yue T. Dispersion of
nonaqueous Co2 Z ferrite powders with titanate coupling
agent and poly(vinyl butyral). Appl Surf Sci 2005;245:252–9.
[26] Karim SN, Hamed O, Mahdi M. Crystallization, shrinkage and
mechanical characteristics of polypropylene/CaCO3
nanocomposites. J Thermoplastic Composite Mater
2013;26:544–54.
Please cite this article in press as: P. Alves, O. Macêdo, G. Barbosa, et al.. High-density polyethylene/mollusk shell-waste compos-
ites: effects of particle size and coupling agent on morphology, mechanical and thermal properties. J Mater Res Technol. 2019.
https://doi.org/10.1016/j.jmrt.2019.01.008
11
[27] Xiao-Lin X, Qing-Xi L, Robert KYL, Xing-Ping Z, Yiu-Wing M.
Rheological and mechanical properties of PVC/CaCO3
nanocomposites prepared by in situ polymerization. Polymer
2004;45:6665–73.
[28] Jang K-S. Mineral filler effect on the mechanics and flame
retardancy of polycarbonate composites: talc and kaolin.
e-Polymers 2016;16:379–86.
[29] Avella M, Cosco S, Di Lorenzo ML, Di Pace E, Errico ME,
Gentile G. Nucleation activity of nano-sized CaCO3 on
crystallization of isotactic polypropylene, in dependence on
crystal modification, particle shape, and coating. Eur
Polymer J 2006;42:1548–57.
[30] Dong HH, Hong YS. Calcined calcium magnesium acetate as
a superior SO2 sorbent: I. Thermal decomposition. AICHE J
2002;48:2971–7.
[31] Sheng LN, Kui Hua H, Chun Mei L. Kinetic calculations for
the thermal decomposition of calcium propionate under
non-isothermal conditions. Chin Sci. Bull 2011;56:1278–84.
[32] Masaya K, Naoki H, Makoto K, Arimitsu U, Akane O.
Preparation and mechanical properties of polypropylene clay
hybrids. Macromolecules 1997;30:6333–8.
[33] Saminathan K, Selvakumar P, Bhatnagar N. Fracture studies
of polypropylene/nanoclay composite. Part II: failure
mechanism under fracture loads. Polymer Testing
2008;27:453–8.
[34] Manish K, Sharma CS, Pradeep U, Vishal V, Pandev KN, Vijai
K, et al. Calcium carbonate (CaCO3 ) nanoparticle filled
polypropylene: effect of particle surface treatment on
mechanical, thermal, and morphological performance of
composites. J Appl Polymer Sci 2012;124:2649–56.
[35] Wang M, Joseph R, Bonfield W. Hydroxyapatite-polyethylene
composites for bone substitution: effects of ceramic particle
size and morphology. Biomaterials 1998;19:2357–66.
[36] Feng Y, Ashok B, Madhukar K, Zhang J, Zhang J, Reddy KO,
et al. Preparation and characterization of polypropylene
carbonate bio-filler (eggshell powder) composite films. Int J
Polymer Anal Charact 2014;19:637–47.
[37] Essabir H, Bensalah MQ, Rodrigue D, Bouhfid R, Qaiss AK. A
comparison between bio- and mineral calcium carbonate on
the properties of polypropylene composites. Construct Build
Mater 2017;134:549–55.
[38] Bharathiraja G, Jayabal S, Sundaram SK, Rajamuneeswaran S,
Manjunath BH. Mechanical behaviors of rice husk and boiled
egg shell particles impregnated coir-polyester. Macromol
Symp 2016;361:136–40.
[39] Shao-Yun F, Xi-Qiao F, Bernd L, Yiu-Wing M. Effects of
particle size, particle/matrix interface adhesion and particle
loading on mechanical properties of particulate polymer
composites. Composites Part B: Eng 2008;39:933–61.