Composites: Part A 68 (2015) 313–322

Contents lists available at ScienceDirect

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

Synergistic effect of coupling agents and fiber treatments on mechanical

properties and moisture absorption of polypropylene–rice husk
composites and their foam
Shu-Kai Yeh a,⇑, Chia-Chun Hsieh b, Hsiao-Ching Chang b, Christopher C.C. Yen c, Yi-Chun Chang c
a
Department of Materials Science and Engineering, National Taiwan University of Science and Technology, Taipei 106, Taiwan, ROC
b
Department of Chemical Engineering and Biotechnology & Institute of Chemical Engineering, National Taipei University of Technology, Taipei 106, Taiwan, ROC
c
Miniwiz Sustainable Energy Development Co., Ltd., 14F No. 102, Guang Fu South Rd., Taipei 106, Taiwan, ROC

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

Article history: Rice husks and polypropylene were applied as the fibers and matrix, respectively, to make composites.
Received 2 April 2014 Polypropylene-grafted maleic anhydride (PP-g-MA) and styrene ethylene butadiene styrene-grafted
Received in revised form 29 July 2014 maleic anhydride (SEBS-g-MA) were used as coupling agents. The rice husks were also treated with
Accepted 11 October 2014
NaOH, silane, or NaOH + HCl + silane to enhance the effect of the coupling agents. Using a combination
Available online 27 October 2014
of 2 wt% PP-g-MA and 1 wt% SEBS-g-MA, the impact strength of the composite increased, but the tensile
strength and modulus were not reduced relative to the use of PP-g-MA alone. The three treatments –
Keywords:
NaOH, silane and NaOH + HCl + silane – and added coupling agents improved the impact strength and
A. Polymer–matrix composites (PMCs)
B. Fiber/matrix bond
decreased the moisture absorption rate of the composites except for those subjected to the alkaline treat-
D. Mechanical testing ments. The foaming results showed that adding coupling agents improved the cell structure and reduced
E. Surface treatments the density of the foam.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction
NF have many advantages, e.g., low density, low cost, and highly
specific mechanical properties. The modulus of natural fibers can
be as high as 6–80 GPa, comparable to those of glass fibers [1].
However, NF density is significantly lower than that of glass fibers.
The high modulus of NFs comes from the high cellulose content.
Cellulose possesses a Young’s modulus of 140–250 GPa, and it
usually occupies more than 60 wt% of the NFs [2]. Due to their
renewable and biodegradable characteristics, NFs are used to
reinforce composites and this can reduce the impact on the envi-
ronment. The low energy consumption and low carbon footprint
of making natural fibers is considered sustainable.
There are many kinds of NFs, such as bamboo, sisal, jute, hemp,
flax, and rice husks, and their use usually depends on their cost and
availability. In this study, rice husks are used as fillers because they
are abundant and usually considered as agricultural waste in
eastern Asia. Rice is grown as an annual plant, and it can grow to
1–1.8 m tall, depending on the variety and soil fertility. Rice husks
usually contain 35–45 wt% cellulose, 19–25 wt% hemicellulose,
20 wt% lignin, 14–17 wt% wax, and 15–17 wt% silica [3]. Like other
⇑ Corresponding author. Tel.: +886 2 2737 6516; fax: +886 2 2737 6544.
E-mail address: skyeh@mail.ntust.edu.tw (S.-K. Yeh).
NFs, rice husks easily absorb moisture because they contain cellu-
lose. It is noteworthy that silica is not seen in most natural fibers.
The high silica content in rice husks provides high stiffness and
improves the flame retardant characteristics of polymer/rice husk
composites [4,5]. Therefore, the use of rice husks as a filler of com-
posites has gradually increased because this solves the problem of
agricultural waste use and produces value-added composites. This
research is especially important in Asia, where most of the world’s
rice is produced.
Rice is a major grain crop and is the staple food for more than
half of the human population. According to a survey by the Food
and Agriculture Organization of the United Nations (FAO) in
2012, the annual rice production reached 724.5 million tons, of
which 655.1 million tons were produced in Asia [6]. Therefore,
as an agricultural waste, rice husks could cause environmental
problems. The current applications of rice husks are as feedstock,
fertilizer, biodiesel, incineration, and pillow stuffing. The idea of
using rice husks or other agricultural waste as fillers in composites
perhaps proposed by researchers as early as in 1985 [7]. However,
this research was not dynamic until 2000. Before 2000, Fuad et al.
were almost the only research group using rice husk ash but not
rice husks as the filler of composites [8,9]. Although the calorie
value of rice husks is low, it is common practice to burn rice husks
and use them as fuel for industrial purposes. The byproduct, rice

http://dx.doi.org/10.1016/j.compositesa.2014.10.019
1359-835X/Ó 2014 Elsevier Ltd. All rights reserved.
314 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322
husk ash, contains nearly 95 wt% silica [10]. Since the mechanical
properties of silica are significantly higher than those of polymers,
one of the potential applications for rice husk ash is to use it as a
filler to improve the mechanical properties of polymer composites.
These days, the biodegradability of polymer composites is becom-
ing more and more important, and researchers are starting to use
rice husks for composite applications. Yang et al. did a series of
studies comparing the mechanical and thermal properties of com-
posites made of wood flour and rice husks. The results showed that
although the mechanical properties of rice husks are slightly lower
than those of wood flours [11], rice husks provide better thermal
stability than wood flours owing to the presence of silica [12].
Polypropylene (PP) is a thermoplastic polymer with a density of
0.81–0.91 g/cm3, which is lower than that of other plastics such as
acrylonitrile butadiene styrene (ABS), polyethylene (PE), polysty-
rene (PS), polyvinyl chloride (PVC), etc. PP also has higher tensile
strength and modulus and better thermal stability than other plas-
tics. The melting point of industrial-grade PP is 160–166 °C [13].
Thus, the service temperature of PP can be as high as 100–
120 °C. The applications of PP include automotive, appliances, daily
necessities, furniture, packaging, etc. Additionally, PP possesses
excellent chemical resistance. In most cases, PP is non-reactive
and it cannot be dissolved by the majority of solvents.
Owing to the biodegradable nature of biofibers, automobile
manufacturers such as Audi, Ford, and BMW in Europe have intro-
duced natural fibers such as bast fibers and wood flours to replace
glass fibers in cars [14]. Although polypropylene–rice husk com-
posites have been used for applications such as utensils, furniture,
and building and construction materials, the use of rice husks for
automobile applications has not been reported. With proper
design, natural fiber composites (NFCs) could become a fashion
product. For example, recycled PP can be compounded with
35 wt% rice husks and injection molded to form a cellular phone
case, wine bottle packaging, or sun glasses [15]. The concept of
integrating fashion design with green/renewable products may
open up a new route for recycled plastics and natural fibers.
A major problem in making polymer/NF composites is the
incompatibility between the fiber and the polymer matrix. Since
polymers are hydrophobic and natural fibers are hydrophilic, the
compatibility between PP and natural fiber is poor. Compounding
two incompatible substances may cause fibers to agglomerate,
and there is a poor bonding between the polymer matrix and the
fibers that may result in poor mechanical properties and a high
rate of moisture absorption, which could be detrimental to the
composites. Also, biofibers tend to suffer fungal attack under
humid conditions, and absorbing moisture may swell the fibers
and lead to swelling of the composites.
The problem can be solved by using coupling agents. Coupling
agents are block copolymers that react with the hydroxyl groups
on the fiber surface to enhance the compatibility between the fiber
and the polymer. There are two ways to modify the surface
properties of natural fibers. First, coupling agents containing isocy-
anates, silane, polymers grafted with maleic acid, and triazine can
be used as polymer additives during compounding. Secondly,
before compounding, the surface of the fibers can be modified with
chemicals used to enhance compatibility. As compared to using
polymer additives, surface modification would be more efficient
than using a coupling agent. However, it is less cost effective and
more time consuming. Surface modification is different from add-
ing coupling agents. Surface modification is a way of altering the
hydrophilicity of the fiber surface. Fiber treatments include
physical methods and chemical methods [16]. Corona or plasma
treatments are considered physical treatments. Chemical methods
include silane treatment, alkaline treatment, acetylation, maleated
coupling, and enzyme treatment [17]. In the treatments mentioned
above, some of them add functional groups onto the surface of the
fibers, whereas others remove the weak parts of the natural fibers.
For example, alkaline treatment removes the weak parts of the nat-
ural fibers, i.e., hemicellulose and lignin, and thus improving the
mechanical properties of the fibers [18], whereas silane treatments
improve the functionality of the fibers and make them compatible
with polymers. In addition to coupling agents or surface modifica-
tion, the type of compounding equipment or changes to the pro-
cessing variables may also reduce the rate of water absorption
[19,20]. The kinetics of moisture absorption are determined by
two factors. One is the diffusion coefficient and the other is the
equilibrium moisture content. The diffusion coefficient determines
the time needed to reach the equilibrium moisture. The equilib-
rium moisture may affect the final dimensions and mechanical
properties of the composite.
The density of most biofibers, including rice husks, is 1.2–1.5 g/
cm3, which is significantly higher than that of solid wood [1]. To
mimic the density of wood and other natural fibers, it is reasonable
to foam NFCs during processing. It is especially desirable to foam
NFCs for automobile applications since it could significantly
improve the fuel efficiency of cars [14]. The foaming of natural fiber
composites can be achieved by various processing techniques such
as batch, injection molding, extrusion, and compression molding
[21]. In these processes, both physical and chemical blowing agents
are applied. Carbon dioxide is considered a possible candidate as a
physical blowing agent because it possesses relatively higher solu-
bility than other inert gas in most polymers and because it contrib-
utes far less to the greenhouse effect as compared to traditional
blowing agents such as chlorofluorocarbons. In this study, carbon
dioxide was used as the blowing agent in order to foam PP–rice
husk composites using the batch foaming method.
2. Experimental
2.1. Materials
Rice husk flours of 35–40 mesh were provided by Miniwiz
Sustainable Energy Development Co. Ltd., Taiwan. Polypropylene,
PP, (PP TAIRIPRO K1023) was purchased from Formosa Chemicals
& Fibre Corporation, Taiwan. The melt flow index of PP is 25 g/
10 min at 230 °C and 2.16 kg. The coupling agent PP-g-MA (trade
name PolybondÒ 3200) was purchased from Chemtura, USA. The
maleic anhydride content of PP-g-MA is 1 wt%. The SEBS-g-MA
(trade name FG-1901) coupling agent was purchased from Kraton,
USA. The maleic anhydride content of FG-1901 is 1.7 wt%. Before
the materials were compounded, rice husks, PP, and the coupling
agents were dried in a vacuum oven at 80 °C for 12 h. It is worth
noting that the coupling agent content is based on the weight per-
centage of rice husk rather than the total weight of the composite.
The rice husks, PP, and additives were compounded using a
Brabender plasticorder (PLE-331) internal mixer with a speed of
50 rpm at 180 °C for 5 min. After mixing, the materials were
quenched in a water bath. The compounded materials were then
ground into powder using a pulverizer (RT-02A, Rong Tsong Preci-
sion Technology). The ground powders were dried at 85 °Cfor 12 h
and then injection molded using a Thermo Haake Minijet injection
molding machine with an injection temperature of 180 °C and a
pressure of 750 bar (75 MPa) to produce ASTM D638 Type IV sam-
ples and ASTM D256 samples for mechanical testing.
a. Mechanical properties testing
The tensile and impact strength and stiffness properties of these
samples were measured using a Tinius Olsen H5KS universal test-
ing machine and a CEAST digital impact tester. The tensile strength
was analyzed according to ASTM D638, whereas the impact test
was carried out according to ASTM D256. The strain rate of the
 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322
tensile test was 1 mm/min. Because NFCs generally showed rela-
tively low impact strength in the standard Izod impact test, the
reversed notch impact test method was applied to enhance the dif-
ferences resulting from the presence of the coupling agents. Note
that a minimum of five samples were tested in each case.
b. Rice husk surface treatment
2.1.1. NaOH chemical surface modification
The rice husks were soaked in 10 wt% NaOH (pH of 11.86) at
room temperature for 7 h. The rice husk flours were then rinsed
with distilled water and aired.
2.1.2. Silane modification
The silane used in this experiment was 3-aminopropyltrieth-
oxysilane, which was purchased from Sigma–Aldrich. The rice husk
flours were first soaked in 3 wt% silane solution at room tempera-
ture for 3 h. The rice husk flours were then washed and aired.
2.1.3. NaOH, HCl and silane chemical surface modification
The rice husk flours were soaked in 10 wt% NaOH at room tem-
perature for 7 h. The alkaline solution was then neutralized by add-
ing 4 wt% HCl. After 10 min, the rice husk flours were rinsed with
distilled water and aired. Finally, the rice husk flours were soaked
in a 3 wt% silane solution at room temperature for an additional 3 h
and then rinsed with water and aired. The treated and non-treated
rice husks were all dried at 80 °C for 12 h before compounding.
2.2. Foaming
The rice husk composites samples were foamed using a batch
foaming system to investigate the foaming feasibility of PP–rice
NFCs. A schematic diagram of the foaming system is shown in
Fig. 1. In this system, three high-pressure inline filters (Swagelok
SS-2F-05) were used as the foaming cells. The length and diameter
of the cells were 40 and 14.3 mm, respectively, and the thickness of
the cell wall was 2 mm. An ISCO 265D syringe pump was con-
nected to the foaming cells via a four-way union to deliver CO2
at a controlled pressure. Each foaming cell was connected to a
three-way ball valve (Swagelok SS-41XS2) for pressure release.
The foaming cells were laid horizontally and immersed in a water
or oil bath to control the foaming temperature at 170 °C. The
Three-way valve
CO2
Syringe Pump
Fig. 1. Schematic diagram of the CO2 batch foaming process. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
this article.)
315
weight of the solid sample was controlled at 0.05 g in each exper-
iment. Samples were immersed in CO2 for 3 h under controlled
temperature. The CO2 immersion time was selected because
long-term exposure of NFCs to temperatures above 100 °C may
lead to severe degradation and change the appearance of the com-
posite. The foaming process was conducted by quickly opening the
three-way ball valve. The foaming cells were immediately
quenched in ice water to freeze the cell structure. The ice and
water mixture was force convected by agitation to ensure fast cool-
ing. The time gap between the foaming and cooling process was
well controlled to be within 2 s. In this study, the foaming temper-
ature was controlled from 160 °C to 170 °C and the foaming
pressure was 13.79 MPa.
2.3. FTIR analysis
The treated rice husks and the rice husks extracted from the
composites were analyzed using FTIR. In the first case, 5 mg of
treated rice husks were mixed with potassium bromide (KBr),
compressed into disks, and then analyzed with the help of a
Perkin–Elmer Spectrum GX system FTIR. In the second case, the
polymer matrix of the composite was dissolved in boiling xylene
and the rice husks extracted from the NFCs were analyzed. In both
cases, the rice husks and KBr were separately pre-dried in an oven
at 80 °C to ensure that the moisture content was less than 1 wt%.
The xylene was purchased from J.T. Baker, whereas the KBr was
purchased from Fluka.
2.4. Scanning electronic microscopy
The surface morphology of treated rice husk, tensile fractured
surfaces of the composites or foamed composites were sputter
coated with gold and the surface morphology was observed using
a Hitachi S-3000H scanning electronic microscope.
2.5. Moisture absorption tests
The rate of water absorption of the WPCs was determined using
a ‘‘blot and weigh’’ method. Three weighed samples were taken
from each batch to conduct this test. The compounded NFCs were
injection molded into disks for testing. The diameter of each disk
was 34 mm and the thickness was 1 mm. Before the test, the
Union
Cell
Water or oil Bath
316 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322
sample disks were placed on a tray of desiccants and dried in a vac-
uum oven at 80 °C for 12 h. After that, each sample was immersed
in a sealed container full of distilled water at room temperature
and each was periodically removed and weighed. The results of
the water absorption test were plotted as the percentage of weight
gained versus time of immersion. These samples are considered to
have reached the equilibrium moisture when the weight gain was
less than 3 mg within a two-week period. The diffusion coefficients
of the NFCs were calculated using the following equations [22]:
!12
Mt Dt
¼4
M1 pð2lÞ2
where M1 is the equilibrium increase in the sample mass (mg) and
2l is the sample thickness (mm). The diffusion coefficient (m2/s) can
1=2
be calculated from the initial slope of MM1t versus t 2l using the short-
time water uptake data.
3. Results and discussion
3.1. Thermogravimetric analysis
Ten milligrams of rice husks were tested in nitrogen at a heating
rate of 10 °C/min. The thermogravimetric analysis (TGA) results for
the rice husks are shown in Fig. 2. The weight decrease from 0 to
100 °C was due to the removal of moisture. Also, the maximum
degradation rate occurred at 300 °C, which indicates a-cellulose
decomposition. After the TGA experiment, 30 wt% of residue
remained. This was the rice husk ash, whose major component is
silica. The results are consistent with those in literature.
3.2. FTIR analysis of treated and untreated fibers
The FTIR spectra of the chemically treated and untreated rice
husks are shown in Fig. 3. Both treated and untreated rice husks
show significant differences in the spectrum of 3200–3400 cm1,
which represents the stretching vibration of intermolecular hydro-
gen-bonded –OH groups in the cellulose fibers [23–25]. Also, a
broad valley was found at 1020–1050 cm1. This could be the
peak of cellulose and SiO2 [18,25]. The evidence of silica reacting
with NaOH is shown at the peak of 786 ± 5 cm1. After the alkaline
treatment, the peak at 786 cm1 decreased or disappeared. It is
possible that part of the cellulose protected by the silicon–cellulose
membrane on the rice husk was degraded during the silica
removal, which may explain the lack of increase of the –OH group
bonds after NaOH treatment [23,25]. SiO2 particle dissolution in
the rice husks created holes in the husks and affected the moisture
absorption behavior of the composites. The results of moisture
Fig. 2. TGA analysis of rice husks.
ð1Þ
Fig. 3. FTIR spectra of untreated and chemically treated rice husks. (For interpre-
tation of the references to colour in this figure legend, the reader is referred to the
web version of this article.)
absorption will be discussed later in this paper. The peaks at
1600 and 1510 cm1 represent the bonds of the aromatic rings of
lignin. Since the peak at 1600 cm1 also represents the structure
of carbonyl groups, the only peak that can be used to estimate lig-
nin removal is at 1510 cm1. The FTIR spectra of the pristine rice
husks and the silane-treated rice husks showed a significant peak
at 1510 cm1. Further, the lignin peak of the alkaline treated rice
husks disappeared, which indicates that most of the lignin was
removed after alkaline treatment. The peak at 1740 cm1 appeared
on the untreated rice husks and either represents the vibration of
the carbonyl groups of the hemicellulose [18] or it is due to the
wax or natural fat in the rice husks [23]. After alkaline treatment,
the 1740 cm1 peak disappeared, indicating the removal of hemi-
cellulose. Since 3-aminopropyltriethoxysilane is known to react
with carbonyl groups and form urethane bonds, the carbon–
oxygen double-bond peak of the silane-treated rice husk also
disappeared [26].
3.3. Mechanical properties analysis
The tensile and impact properties of NFCs containing 0–50 wt%
rice husks are shown in Fig. 4(a) and (b), respectively. The modulus
increased with the rice husk loading level, whereas the strength
decreased. Also, the impact strength decreased with the loading
level of rice husks. Both phenomena were expected and have been
reported previously [27]. The effect of coupling agents was pro-
nounced. When 1 wt% of PP-g-MA or SEBS-g-MA was added to
the composite, the composite strength increased by at least 18%.
The results are shown in Fig. 5(a) and (b) for PP-g-MA and SEBS-
g-MA, respectively. The PP-g-MA addition seemed to have little
impact on the modulus of the composite, which is another major
concern of NFCs. The composite modulus remained unchanged in
the presence of PP-g-MA. Indeed, composite modulus decreased
with the loading level of SEBS-g-MA. The results were expected
since the modulus of rubber is usually much lower than that of
polymer or fibers. Also, the strength of the composites started to
decrease after the content of SEBS-g-MA exceeded 2 wt%. Thus,
in terms of tensile strength and modulus, PP-g-MA is a more effec-
tive coupling agent than SEBS-g-MA.
The impact strength of the composites containing different
amounts of SEBS-g-MA or PP-g-MA is shown in Fig. 5(c). The
results show that adding SEBS-g-MA provided better improvement
in impact strength than PP-g-MA. Although adding SEBS-g-MA
improved the impact strength of the composite, it had other
 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322 317

40 2500 90

Tensile Strength (MPa)

Impact Strength (J/m)

35

Modulus (MPa)
2000 80
30
25 70
1500
20 60
15 1000
50
10 Tensile Strength 500
5 40
Modulus
0 0 30
0 10 20 30 40 50 10 20 30 40 50
Rice Husk (wt%) Rice Husk (wt%)
(a) (b)
Fig. 4. (a) Tensile properties and (b) impact strength of NFCs containing different amounts of rice husks.

Young's Modulus (MPa)

Young's Modulus (MPa)
Tensile Strength(MPa)

35 2400

Tensile Strength(MPa)
35 2400
31 2200 31 Tensile Strength 2200
Modulus
27 2000 27 2000

23 1800 23 1800
Tensile Strength
19 1600 19 1600
Modulus
15 1400 15 1400
0 1 2 3 4 5 0 1 2 3 4 5
PP-g-MA Content (wt%) SEBS-g-MA Content (wt%)
(a) (b)
Izod impact strength(J/m)

70
65
PP-g-MA
60
SEBS-g-MA
55
50
45
40
35
30
0 1 2 3 4 5
Coupling Agent Content (wt%)
(c)
Fig. 5. Tensile properties of NFCs containing 50 wt% rice husk and different amounts of (a) PP-g-MA and (b) SEBS-g-MA. (c) Impact strength of NFCs with different amounts of
PP-g-MA or SEBS-g-MA added.

drawbacks. Adding SEBS-g-MA reduced the modulus of the 3.4. Microstructure analysis of tensile fractured sample
composite. In the presence of only 2 wt% of SEBS-g-MA in the com-
posite, composite modulus decreased by 11%, which is undesirable. SEM images of tensile fractured NFCs containing different
Besides, PP-g-MA helped improve the impact strength of the coupling agents are shown in Fig. 7(a)–(d). The fractured sur-
composite without decreasing the modulus. Consequently, it is face of the composite without a coupling agent is shown in
reasonable to use both PP-g-MA and SEBS-g-MA as the coupling Fig. 7(a): it shows many fiber pull-out holes, which indicate
agent for the composite. poor bonding between the fiber and polymer matrix. After
Since the composite has to meet minimum strength require- 2 wt% of PP-g-MA was added, holes disappeared and fractured
ments for specific applications, the content of PP-g-MA was fixed fibers were found, as shown in Fig. 7(b). The addition of a cou-
at 2 wt% and that of SEBS-g-MA was varied. The tensile and impact pling agent clearly improved the adhesion between the fiber
properties of NFCs containing 2 wt% PP-g-MA plus various and the matrix. When 2 wt% of SEBS-g-MA was added into
amounts of SEBS-g-MA are shown in Fig. 6(a) and (b), respectively. the composite, the morphology of the fractured surface
The results show a clear synergistic effect between the two cou- changed. In Fig. 7(c), ductile failure behavior is observed in
pling agents. In addition to 2 wt% PP-g-MA, adding 1 wt% SEBS-g- the presence of 2 wt% of SEBS-g-MA. However, as can be seen
MA into the composite improved the tensile and impact strength in Fig. 7(d), when 2 wt% of PP-g-MA was added together with
of the composite, and the modulus remained intact. The tensile 1 wt% of SEBS-g-MA, the fractured surface showed much less
strength of the composite increased from 17.8 to 25.6 MPa, fiber pull outs and less polymer laceration. This could explain
whereas the impact strength increased from 36 to 50 J/m in the why adding 1 wt% of SEBS-g-MA plus 2 wt% of PP-g-MA had
presence of 2 wt% PP-g-MA plus 1 wt% SEBS-g-MA. When the very little effect on the modulus of the composite. The micro-
loading level of SEBS-g-MA exceeded 1 wt%, the strength of the scopic images of the fractured surfaces can be correlated with
composite started to decrease and impact strength did not increase the mechanical properties data and help explain the effect of
further. Therefore, 2 wt% PP-g-MA with 1 wt% SEBS-g-MA can be different coupling agents. Similar results were observed in
considered the optimized coupling agent loading level. wood–plastic composites [28].
318 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322

27 2300 70

Young's Modulus (MPa)

Tensile Strength(MPa)

Impact Strength (J/m)

2100 60
24
1900 50
21
1700 40

18 1500 30
0 1 2 3 4 5 0 1 2 3 4 5
SEBS-g-MA Content (wt%) SEBS-g-MA Content (wt%)
(a) (b)
Fig. 6. (a) Tensile properties and (b) impact properties of NFCs containing 50 wt% rice husks and 2 wt% PP-g-MA plus different amounts of SEBS-g-MA. The x-axis is the
loading level of SEBS-g-MA.

(a) (b)

(c) (d)
Fig. 7. Fractured surface of NFCs containing 50 wt% rice husks with (a) no coupling agent, (b) 2 wt% PP-g-MA, (c) 2 wt% SEBS-g-MA, and (d) 2 wt% PP-g-MA + 1 wt% SEBS-g-
MA.

The FTIR spectra of fibers extracted from the composites are

shown in Fig. 8. As can be seen from the figure, the hydroxyl group
peak at 3200–3400 cm1 decreased significantly when coupling
agents were added. This indicates the maleic anhydride reacting
with the hydroxyl groups on the fibers, thus improving the NFC
mechanical properties.
For untreated fibers, the experiment results showed that adding
a coupling agent could significantly improve the tensile and impact
strength of the composites. It is of interest to know whether fiber
surface treatment could further improve the strength of the com-
posites. Thus, the surface-modified fibers were compounded with
PP and the mechanical properties of the composites were measured.
The mechanical properties of NFCs containing 50 wt% treated or
untreated rice husks and different coupling agents are listed in
Table 1. The results show that no matter what kind of treatment
was applied to the fibers, adding coupling agents improved the
strength of the composite from 17–18 MPa to 23–27 MPa; however,
the fiber treatments may or may not improve the tensile strength of Fig. 8. FTIR spectra of untreated rice husks extracted from composites with
the composites. Alkaline treatments are known to increase the ten- different coupling agents added. (For interpretation of the references to colour in
sile strength of rice husk composites since it removes the fat, waxes, this figure legend, the reader is referred to the web version of this article.)
 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322 319

Table 1
Mechanical properties of PP–rice husk composites containing different coupling agents and different fiber treatments. The rice husk content is 50 wt%.

Modulus Standard Tensile strength Standard Impact strength Standard

(MPa) deviation (MPa) deviation (J/m) deviation
Untreated 2134 51 17.76 0.93 35.83 2.19
Untreated + 2%PP-g-MA 2135 80 24.46 0.59 42.03 3.10
Untreated + 2%PP-gMA + 1%SEBS-g-MA 2214 24 25.6 0.75 50.42 1.51
NaOH 2124 39 18.68 0.77 43.02 2.78
NaOH + 2%PP-g-MA 2362 28 22.47 1.18 49.50 2.92
NaOH + 2%PP-g-MA + 1%SEBS-g-MA 2273 58 23.07 0.58 55.34 3.79
Silane 2110 38 17.74 1.34 41.25 2.06
Silane + 2%PP-g-MA 2284 84 21.92 1.00 47.44 4.12
Silane + 2%PP-g-MA + 1%SEBS-g-MA 2184 73 23.50 0.54 54.31 4.82
NaOH + HCl + silane 2107 63 18.47 1.05 42.72 1.01
NaOH + HCl + silane + 2%PP-g-MA 2343 89 25.98 0.90 50.87 2.50
NaOH + HCl + silane + 2%PP-g-MA + 1%SEBS-g-MA 2312 66 27.52 1.51 61.18 4.06

and other impurities in the rice husks. Because of that, more reactive
hydroxyl groups in the fibers would be exposed and the strength of
the composite would be increased [23]. However, this phenomenon
was not observed in this study. In our case, the alkaline and silane
treatments slightly decreased the tensile strength, but the
NaOH + HCl + silane treatment somehow improved the tensile
strength of the composite. Fiber treatments and adding coupling
agents did not affect the tensile modulus and strength of the com-
posites. The modulus of the composites fell between 2100 and
2300 MPa. The fiber treatments had a significant effect to the impact
strength of the composite. As can be seen in Table 1, even without
the coupling agents, the impact strength of the NFCs improved from
36 J/m to 42 J/m. When 2 wt% of PP-g-MA and 1 wt% of SEBS-g-MA
were added, the impact strength of the composites made with
untreated or treated fibers both increased, but the composites made
with treated rice husks increased more. The NaOH + HCl + silane
treatment can improve the impact strength of the composite from
50 J/m to almost 61 J/m. In summary, coupling agents in conjunc-
tion with fiber treatments helped increase the impact strength of
the NFCs from 36 J/m to 61 J/m, which is an increase of 70%. Fiber
treatments may not be helpful in increasing the tensile properties,
but they definitely make a difference in improving the impact prop-
erties of the composite.
The results might be explained by observing the surface of the
treated fibers. The surfaces of the treated and untreated rice husks
are shown in Fig. 9(a)–(d). The surface morphology of the untreated
rice husks and the silane-treated rice husks is shown in Fig. 9(a) and
(b), respectively. In both figures, the fiber surfaces remained intact.
When the fibers were treated with silane, the surface of the fibers
seemed to be covered with a layer of silane. The surface morphology
of fibers subjected to alkaline treatment is shown in Fig. 9(c) and (d).
The fibers were broken and many holes were observed on the fiber
surface. It is well known that alkaline will react with silica. Alkaline
treatment will dissolve the SiO2 on the surface of the fibers. The SEM
images provide clear evidence of silica removal and the observation
is consistent with Ndazi et al. [23]. Rice husks contain a significant
amount of silica (rice husk ash) and the impact strength of PP–rice
husk ash composites decreases dramatically with increasing filler
loading [9]. Therefore, removing silica from the composite may
improve the impact strength of the composite. This could be the rea-
son why alkaline-treated fibers result in improved impact strength.
Further, the composite made with silane-treated fibers showed
improvement in impact strength but not tensile strength, which is
not surprising. Bikiaris et al. observed similar phenomena in PP–
glass fiber composites [29]. The glass fibers were treated with 3-
aminopropyltriethoxysilane and PP-g-MA was used as the coupling
agent, which is exactly the same as in this study. Their results
showed that the impact strength of the composite increased signif-
icantly but the tensile strength remained unchanged. It would
appear that the silane used in the current study was not very effi-
cient at transferring stress but it was effective in consuming energy.
Therefore, the impact strength of the NFCs improved. In the third
case, both alkaline and silane treatments were applied to the fibers;
a synergistic effect was observed and the impact strength of the
NFCs was further elevated.
3.5. Moisture absorption behavior
Increasing the loading level of natural fiber in a composite is
desirable because it helps reduce the cost and increases composite
modulus since most natural fibers are less expensive and stiffer
than polymers. However, it may not be suitable for outdoor appli-
cations in which moisture absorption is always a concern. Moisture
exposure leads to fungal attack and dimensional instability. More-
over, cyclic moisture absorption/desorption may accelerate the
aging of the composite and decrease its service life. Thus, the mois-
ture absorption behavior and kinetics are of interest.
The diffusion coefficient and equilibrium moisture of the com-
posites made by different coupling agents and surface treatments
are listed in Table 2. The data was analyzed by a two-way analysis
of variance (ANOVA) using Microsoft Excel with a confidence inter-
val of 95%. The analysis results showed that both fiber treatments
and coupling agents were statistically significant in reducing the
diffusion coefficient and the equilibrium content of the composites.
Also, there were interactions between fiber treatments and coupling
agents in terms of the diffusion coefficient of the composites but the
two factors had no interaction with the equilibrium moisture
content. Among the four fiber treatments, the silane treatments
were most effective. Both the silane treatments and the
NaOH + HCl + silane treatments decreased the diffusion coefficient
of the composites by 60–70% in the absence of coupling agents.
Treating fibers with alkaline did not help reduce the diffusion coef-
ficient, and the diffusion coefficient of the composites increased by
50–60% as compared to untreated fibers. This is not surprising since
alkaline treatment removed most of the silica in the rice husk and
created pores in the fibers for moisture diffusion. Moreover, adding
coupling agents further reduced the diffusion coefficients by
another 20–40%, irrespective of the fiber treatment type employed.
Adding a coupling agent only slightly decreased the equilibrium
moisture content of composites with the same fiber treatment. The
likely reason for this is that rice husks are a porous medium, and
99% of the hydroxyl groups lie inside the pores. The size of these
pores ranges from 40 Å to 40 lm, and there is not enough time for
PP-g-MA or SEBS-g-MA to react with the hydroxyl groups inside
these pores during compounding. In other words, coupling agents
react only with the surface hydroxyl groups. However, water mol-
ecules were able to access the interior hydroxyl groups during the
long-term exposure in the absorption test, and so the equilibrium
320 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322

(a) (b)

(c) (d)
Fig. 9. Surface morphology of fibers (a) as received, (b) treated with silane, (c) treated with alkaline, and (d) treated with NaOH + HCl + silane.

Table 2
Diffusion coefficients and equilibrium moisture content of PP–rice husk composites containing different coupling agents and different fiber
treatments. The rice husk content is 50 wt%.

Equilibrium moisture (%) Diffusion coefficient (m2/s)

Untreated 11.39 2.37  1013
Untreated + 2%PP-g-MA 9.91 1.0  1013
Untreated + 2%PP-gMA + 1%SEBS-g-MA 9.38 8.37  1014
NaOH 15.42 2.21  1013
NaOH + 2%PP-g-MA 15.31 1.50  1013
NaOH + 2%PP-g-MA + 1%SEBS-g-MA 14.35 1.33  1013
Silane 10.69 7.08  1014
Silane + 2%PP-g-MA 9.45 4.65  1014
Silane + 2%PP-g-MA + 1%SEBS-g-MA 9.71 5.65  1014
NaOH + HCl + silane 12.5 9.54  1014
NaOH + HCl + silane + 2%PP-g-MA 11.25 4.77  1014
NaOH + HCl + silane + 2%PP-g-MA + 1%SEBS-g-MA 11.32 4.21  1014

moisture content showed only a slight decrease in the presence of
coupling agents [19]. Alkaline treatment increased the equilibrium
content of the composite significantly. The results can be explained
by the SEM images of Fig. 9 and by the FTIR analysis. The silica par-
ticles were dissolved by the alkaline solution, which created many
tiny pores in the rice husks. Further, hemicellulose and lignin were
removed after the alkaline treatment, which resulted in an increase
in the number of hydroxyl groups on the cellulose surface. The
same phenomena have been observed in wood–plastic composites.
Both the diffusion coefficient and the equilibrium moisture content
increased significantly after the fibers were treated with alkaline
[30]. Thus, alkaline treatment tends to worsen the moisture
absorption properties of NFCs. Further, silane or NaOH + HCl + si-
lane treatments seem to have little impact in terms of altering
the equilibrium moisture content. Among all the samples, those
made with fibers treated with NaOH + HCl + silane and 2 wt% PP-
g-MA plus 1 wt% SEBS-g-MA showed the lowest diffusion coeffi-
cient and the highest tensile and impact strength. The results
clearly show that both fiber treatments and coupling agents are
effective in improving the properties of the composites and that
there are clear synergistic effects on both fiber treatments and cou-
pling agents. Adding coupling agents to composites made with sur-
face-treated fibers further decreased the diffusion coefficient and
increased the strength of the composite.
3.6. Foaming of PP–rice husk composites
The cell morphology of foamed rice husk composites containing
different coupling agents is shown in Fig. 10. Since the different
fiber treatments showed very little impact on the cell morphology,
only foamed composites without fiber treatments are shown to
avoid redundancy. The results show that adding a coupling agent
improved the cell structure significantly. Although the cells were
torn by rice husk fibers and the average cell density and cell size
could not be reasonably estimated, the cell size decreased and cell
density increased significantly in the presence of the coupling
agents. Further, the bulk density of the foam decreased. The bulk
densities of samples containing no coupling agent, 2 wt%
 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322
(a)
(c)
Fig. 10. Cell morphology of composite foams containing 50 wt% of rice husks and (a) no coupling agent, (b) 2 wt% of PP-g-MA, (c) 2 wt% of SEBS-g-MA, and (d) 2 wt% of PP-g-
MA plus 1 wt% of SEBS-g-MA. The bulk densities of samples (a)–(d) are 0.71, 0.67, 0.60, and 0.68 g/cm3, respectively.
PP-g-MA, 2 wt% SEBS-g-MA, and 2 wt% PP-g-MA plus 1 wt% SEBS-
g-MA are 0.71, 0.67, 0.60, and 0.68 g/cm3, respectively. The results
are consistent with the result of Park et al. [31].
4. Conclusions
In the past, rice husks were considered agricultural waste and
were used with low-value applications. However, rice husks can
be used as the filler in NFCs and become a value-added product.
In this study, PP–rice husk composites were made by melt com-
pounding and injection molding. The results showed that adding
rice husks as fillers for the composites deteriorates the impact
strength of PP significantly. However, the problems can be
overcome by adding coupling agents such as PP-g-MA and SEBS-
g-MA. Our results showed that simultaneously adding 2 wt% PP-
g-MA and 1 wt% SEBS-g-MA resulted in a synergistic effect. The
impact strength and tensile strength of the composite increased
by 35% and 41%, respectively, whereas the modulus of the NFCs
remained intact in the presence of 1 wt% SEBS-g-MA. In addition
to coupling agents, different fiber treatments such as alkaline and
silane treatments were applied to improve the mechanical proper-
ties of NFCs. Chemical treatment of fibers did not significantly
improve the tensile strength of the NFCs but it helped improve
the impact strength of the composite. In our case, using a coupling
agent and alkaline and silane treatments simultaneously resulted
in a 70% increase in impact strength as compared to NFCs made
without coupling agents and fiber treatments. The results of mois-
ture absorption tests showed that both PP-g-MA and SEBS-g-MA
are effective in reducing the moisture diffusion rate and equilib-
rium moisture content. However, treating fibers with an alkaline
substance increases both the moisture content and rate of
moisture diffusion, both of which are highly undesirable. Compos-
ites made with fibers treated with NaOH + HCl + silane and 2 wt%
PP-g-MA with 1 wt% SEBS-g-MA possessed the lowest moisture
diffusion rate and the highest strength among all the samples.
321
(b)
(d)
The results clearly showed that both fiber treatments and coupling
agents are effective in improving the properties of the composites
and that there is a synergistic effect among fiber treatments and
coupling agents. Finally, the foaming results showed that adding
coupling agents is helpful in reducing the density of the composite
foam and the cell morphology improved significantly in the
presence of coupling agents.
Acknowledgments
This work was funded by contract number NSC 102-2622-E-
027-012-CC3 and by contract number NSC 98-2218-E-006-246
from the National Science Council, Taiwan. The author would like
to thank Mr. Arthur Huang, the founder and CEO of Miniwiz for
supporting his employees joining this project.
References
[1] Dittenber DB, GangaRao HVS. Critical review of recent publications on use of
natural composites in infrastructure. Compos Part A 2012;43(8):1419–29.
[2] Bledzki AK, Gassan J. Composites reinforced with cellulose based fibres. Prog
Polym Sci 1999;24(2):221–74.
[3] Faruk O, Bledzki AK, Fink H-P, Sain M. Biocomposites reinforced with natural
fibers: 2000–2010. Prog Polym Sci 2012;37(11):1552–96.
[4] Kuboki T, Lee YH, Park CB, Sain M. Effects of styrene-ethylene-butylene-
styrene based additives on the mechanical properties of rice hull/
polypropylene composites. Polym Eng Sci 2007;47(7):1148–55.
[5] Zhao Q, Zhang B, Quan H, Yam RCM, Yuen RKK, Li RKY. Flame retardancy of rice
husk-filled high-density polyethylene ecocomposites. Compos Sci Technol
2009;69(15–16):2675–81.
[6] Trade and Markets Division Food and Agriculture Organization of the United
Nations. FAO rice market monitor – July 2012. 2012.
[7] Sanadi AR, Prasad SV, Rohatgi PK. Natural fibres and agro-wastes as fillers and
reinforcements in polymer composites. J Sci Ind Res 1985;44(8):437–42.
[8] Fuad AMY, Ishak MZA, Omar MAK. Rice husk ash as filler in polypropylene:
effect of wax and silane coupling agents. Plast Rubber Compos Process Appl
1994;21(4):225–35.
[9] Fuad MYA, Ismail Z, Ishak ZAM, Omar AKM. Application of rice husk ash as
fillers in polypropylene: effect of titanate, zirconate and silane coupling agents.
Eur Polym J 1995;31(9):885–93.
322 S.-K. Yeh et al. / Composites: Part A 68 (2015) 313–322
[10] Chand N, Dan TK, Verma S, Rohatgi PK. Rice husk ash filled—polyester resin
composites. J Mater Sci Lett 1987;6(6):733–5.
[11] Yang HS, Wolcott MP, Kim HS, Kim S, Kim HJ. Effect of different
compatibilizing agents on the mechanical properties of lignocellulosic
material filled polyethylene bio-composites. Compos Struct
2007;79(3):369–75.
[12] Kim HS, Yang HS, Kim HJ, Lee BJ, Hwang TS. Thermal properties of agro-flour-
filled biodegradable polymer bio-composites. J Therm Anal Calorim
2005;81(2):299–306.
[13] Maier C, Calafut T. Polypropylene: the definitive user’s guide and
databook. Norwich: William Andrew; 1998.
[14] Bledzki AK, Faruk O, Sperber VE. Cars from bio-fibres. Macromol Mater Eng
2006;291(5):449–57.
[15] Yeh S-K, Yen CCC, Bodecker J. Recent progress in polymer natural fibre
composites made from recycled plastics. In: Thakur VK, Singha AS,
editors. Biomass based biocomposites. Shrewsbury: Smithers Rapra; 2013.
p. 115–44.
[16] George J, Sreekala MS, Thomas S. A review on interface modification and
characterization of natural fiber reinforced plastic composites. Polym Eng Sci
2001;41(9):1471–85.
[17] Li X, Tabil LG, Panigrahi S. Chemical treatments of natural fiber for use in
natural fiber-reinforced composites: a review. J Polym Environ
2007;15(1):25–33.
[18] Chang WP, Kim KJ, Gupta RK. Ultrasound-assisted surface-modification of
wood particulates for improved wood/plastic composites. Compos Interfaces
2009;16(7–9):687–709.
[19] Yeh SK, Gupta RK. Improved wood–plastic composites through better
processing. Compos Part A 2008;39(11):1694–9.
[20] Bledzki AK, Letman M, Viksne A, Rence L. A comparison of compounding
processes and wood type for wood fibre – PP composites. Compos Part A
2005;36(6):789–97.
[21] Faruk O, Bledzki AK, Nicolais L. Wood plastic composite: present and future.
Wiley encyclopedia of composites. New York: John Wiley & Sons, Inc.; 2011.
[22] Crank J. The mathematics of diffusion. 2nd ed. Oxford: Clarendon Press; 1975.
[23] Ndazi BS, Karlsson S, Tesha JV, Nyahumwa CW. Chemical and physical
modifications of rice husks for use as composite panels. Compos Part A
2007;38(3):925–35.
[24] Petrarch Systems Inc.. Silicon compounds: register and
review. Bristol: Petrarch Systems; 1987.
[25] Efimov AM. Optical constants of inorganic glasses. Boca Raton: CRC Press;
1995.
[26] Li C, Wilkes GL. The mechanism for 3-aminopropyltriethoxysilane to
strengthen the interface of polycarbonate substrates with hybrid organic–
inorganic sol–gel coatings. J Inorg Organomet Polym 1998;8(1):33–45.
[27] Yeh SK, Casuccio ME, Al-Mulla A. Polypropylene-wood plastic composites
reinforced with different amounts of wood and organoclay. SPE ANTEC 2006.
Charlotte, NC; 2006. p. 209–13.
[28] Yeh S-K, Kim K-J, Gupta RK. Synergistic effect of coupling agents on
polypropylene-based wood–plastic composites. J Appl Polym Sci
2013;127(2):1047–53.
[29] Bikiaris D, Matzinos P, Larena A, Flaris V, Panayiotou C. Use of silane agents and
poly(propylene-g-maleic anhydride) copolymer as adhesion promoters in
glass fiber/polypropylene composites. J Appl Polym Sci 2001;81(3):701–9.
[30] Chang WP, Kim KJ, Gupta RK. Moisture absorption behavior of wood/plastic
composites made with ultrasound-assisted alkali-treated wood particulates.
Compos Interfaces 2009;16(7–9):937–51.
[31] Zhang H, Rizvi GM, Park CB. Development of an extrusion system for
producing fine-celled HDPE/wood–fiber composite foams using CO2 as a
blowing agent. Adv Polym Technol 2004;23(4):263–76.