Characterization and preparation of conductive exfoliated graphene nanoplatelets kenaf fibre hybrid polypropylene composites

Synthetic Metals 212 (2016) 91–104

Contents lists available at ScienceDirect

Synthetic Metals journal homepage: www.elsevier.com/locate/synmet

Characterization and preparation of conductive exfoliated graphene nanoplatelets kenaf fibre hybrid polypropylene composites Christopher Igwe Idumaha,b , Azman Hassana,* a Enhanced Polymer Research Group (EnPRO), Department of Polymer Engineering, Faculty of Chemical Engineering, Universiti Teknologi Malaysia, 81310 Skudai, Johor Darul Takzim, Malaysia b Ebonyi state university, Nigeria

A R T I C L E I N F O

A B S T R A C T

Article history: Received 22 September 2015 Received in revised form 8 December 2015 Accepted 11 December 2015 Available online xxx

Exfoliated graphene nanoplatelets (GNP) kenaf fiber (KF) hybrid polypropylene (PP) composites materials were prepared by melt extrusion followed by injection molding. PP/KF/MAPP/GNP composites 0–5 phr were prepared and characterized using X-ray diffraction (XRD), differential scanning calorimetre (DSC), thermogravimetry analysis (TGA), heat deflection temperature (HDT), thermal mechanical analysis (TMA), Fourier transform infrared (FTIR) spectroscopy analysis, field emission scanning electron microscopy. The morphological studies revealed a homogenous dispersion of GNPs in PP/KF/MAPP/GNP up to 3 phr loading after which agglomeration occurred. Flexural strength and modulus were enhanced by 70% and 98% respectively at 3 phr GNPs loading which were the highest values obtained. Interestingly, the highest value for the impact strength was also recorded at 3 phr loading. Thermal conductivity increased by 88%, CTE decreased by 80%, water absorption and thickness swelling decreased while HDT improved. The thermal stability of the composites were generally improved at all GNP loading with the highest at 3 phr. From the overall results, it is obvious that the optimum concentration of GNPs in the PP/ KF/MAPP/GNP system in terms of both mechanical and thermal properties was 3 phr loading. Although, the mechanical and thermal properties of the composites were improved, the FTIR analysis did not reveal any chemical interaction between GNP and the PP/KF/MAPP system. ã 2015 Published by Elsevier B.V.

Keywords: Polymer nanocomposites GNP thermal conductivity coefficient of thermal expansion heat deflection temperature mechanical

1. Introduction Since the isolation of graphene in 2004, there has been escalating interest in studies with graphene inclusion. Many research investigations have been conducted and numerous studies are emerging in various aspects of its unique properties. Nowadays, research investigations into the structure and properties of graphene have advanced from inquisitive based to applicability [1]. Graphene is a single-layer carbon nanoparticle composed of sp2 hybrid carbon atoms aligned hexagonally in planar structures. Several unique properties have enhanced the suitabilty of this material for diverse applications such as its exceptionally high mechanical strength with Young’s modulus of 1 TPa, and tensile strength of 20 GPa [2,3], excellent electrical (5,000 S/m) and thermal conductivities at 3000 W/m.K. Research investigations using this material has escalated in numerous areas such as conductive polymer composites (CPCs) and intrinsically conducting polymers (ICPs) for wide range of engineering

* Corresponding author. E-mail address: [email protected] (A. Hassan). http://dx.doi.org/10.1016/j.synthmet.2015.12.011 0379-6779/ ã 2015 Published by Elsevier B.V.

applications such as thermal interface materials, bio-actuators, fuel cells, drug delivery, tissue engineering, antennas, neuralprobes, biosensors, chemical sensors and so on [1,4–8]. Nowadays, exfoliated graphene nanoplatelets (GNP) have emerged as a new reinforcing filler for the enhancement of mechanical properties [9–11], thermal properties [12,13], and barrier properties [14] of GNP reinforced polymeric nanomaterials. A major factor for escalating application of GNP in production of polymer nanocomposites is the lower cost of graphene, which is the precursor for GNP [15]. Results from various researches have also demonstrated that GNPs exhibited excellent conductivity and reduced the percolation threshold of nanocomposites. In addition, advantages shown by GNPs when compared with other types of nanofillers such as carbon nanotubes (CNTs), include lowered cost, layered structure similar to nanoclays for ease of dispersion and processability during composite preparation [9]. Several techniques used in composite preparation include melt mixing, in situ exfoliation, and solution polymerization. From an industrial perspective, melt polymerization has proven the most convenient type of processing technique because of easy adaptation to commonly available plastic processing machines such as extruders, in addition to being more economical as zero solvents

92

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

Table 1 BET parameters. Vm

=

Calculated value from 1/(slope + intercept) in graph of 1/(Va (Po/P) 1) versus P/Po

N a m 22, 400

= = =

Avogadro constant given as 6.0  1023 mol1 Cross sectional area of single adsorbate molecule in meters square (0.195 nm2 for krypton and 0.162 nm2 for nitrogen) Mass of GNP test flour in grams Volume of 1 mol of adsorbate gas at STP allowing for slight shift from the ideal, in milliliters

are involved which may also impact negatively on the environment. Numerous studies have been conducted using direct melt mixing technique for even distribution of graphene in polymer matrix [13–17,25]. Polypropylene (PP) is a light weight, low cost, linear-oleifinic commodity thermoplastic with good processability applied in packaging and fibre production. However, PP is characterized by low stiffness and flexural modulus, and poor thermal properties which positions it as a poor material where these properties are essential [25,26]. Numerous investigations have been conducted to enhance the mechanical, thermal, and water absorption properties of biocomposites through inclusion of nanoparticles [14–16]. In a previous study [16], the effects of nanographene on the physico-mechanical properties of bagasse/ polypropylene composites was investigated. Results revealed that inclusion of GNP enhanced overall physico-mechanical properties of the material. In another study, multilayered graphene efficiently formed by mechanical exfoliation for nonlinear saturable absorbers in fiber mode-locked lasers [20]. Results revealed improvement of material properties as a result of the inclusion of multilayered graphene. Another study reported effect of GNP on physical properties of polactic acid kenaf fibre filled GNP [13]. Other studies using nanoparticles for enhancement of biocomposite material properties have been reported [29,30]. Results also indicated enhancement of properties through GNP inclusion. To the best of our knowledge, thermal conductivity, coefficient of thermal expansion, morphological characteristics and properties of PP/ KF/MAPP filled with GNPs have not received any attention in the open literature. Therefore, in the present study, GNP filled PP/KF/ MAPP/GNP composites were developed and their thermal conductivity, coefficient of thermal expansion, thermal stability, morphology, heat deflection temperature, impact strength, flexural strength and modulus, water absorption and thickness swelling were investigated. 2. Experimental 2.1. Materials SM 240 grade heterophasic polypropylene (PP) copolymer of melt flow index 35 g/10 min (230
C and 2.16 kg load) and density of 0.901 g/cm3 was purchased from Lotte Titan Chemicals Malaysia. Maleic anhydride grafted polypropylene (MA-g-PP) compatibilizer of melt flow index 150 g/10 min (230
C and 2.16 kg load) and melt temperature 167
C was purchased from Dupont, Dow Elastomers, and Wilmington DE, USA. Kenaf fiber was obtained from Malaysian Agricultural and Development Institute (MARDI), Kuala Lumpur. Exfoliated graphene nanoplatelets, GNP-M5 grade containing 99.5% carbon and graphene nanoplatelets of average diameter 5 mm, and average thickness of 6 nm was purchased as dry flour from XG Sciences, Inc., East Lansing, MI, USA, and applied as

Table 2 Formulation of uncompatibilized PP/GNP nanocomposites. Sample designation

PP (phr)

GNP (phr)

PP/GNP 3

100

3

received. Calculated Brunauer, Emmett and Teller (BET) surface area of specimens applied in this study is 158 m2/g obtained from laboratory measurement. 2.2. BET surface area measurement The Brunauer, Emmett and Teller (BET) surface area was ascertained using Gemini V Surface Area Analyzer of isotherm nitrogen adsorption set at 77 K. Prior to BET evaluation, the specimen was degassed at 623 K for 4 h under atmospheric pressure. Eq. (1) was used in the calculation of specific BET surface area of GNP. Table 1 explains parameters used in Eq. (1). S¼

V m Na 22400m

ð1Þ

2.3. Sample preparation PP was dried in a vacuum oven for 24 h at 80
C, while MA-g-PP was dried for 8 h at 60
C. Kenaf core fibre was ground and sieved using mesh of <500 mm using sieve shaker equipment to obtain kenaf fiber flour of size <500 mm. In order to reduce moisture content, kenaf flour was oven dried at 60
C for 24 h. The composites were melt intercalated in a single step process using Brabender PL 2000 Plastic Coder counter rotating double screw extruder at optimized temperature of 185–20060
C from feed zone to die head zone according to sample formulations shown in Tables 2 and 3 for uncompatibilized PP/GNP-3 phr and PP/KF/ MAPP/GNP 0–5 phr nanocomposites respectively. Extruder screw speed was maintained at 60 rpm. After melt intercalation, the extrudates were subsequently pelletized. In order to eliminate moisture, pellets were oven dried at 8060
C for 24 h, prior to injection molding to standard mechanical tests specimens using JSW model NIOOB 11 Muraron-Japan injection molding machine at temperature range of 185–200
C. As shown in formulation, fibers were varied from 20 wt.%. This is because preliminary experiments revealed fiber control difficult at loadings 30 wt.% due to filler– fibre agglomeration. This also agreed with previous works were fiber content were kept constant at 20 wt.% for equal reason [20]. The concentrations of GNP were calculated based on parts per hundred of total composites (phr).

Table 3 Sample formulation for PP/KF/MAPP/GNP hybrid nanocomposites. Sample designation PP

PP (wt.%) 100

MAPP (wt.%) 0

KENAF (wt.%) 0

GNP (phr) 0

PP/KF/MAPP PP/KF/MAPP/GNP PP/KF/MAPP/GNP PP/KF/MAPP/GNP PP/KF/MAPP/GNP PP/KF/MAPP/GNP

75 75 75 75 75 75

5 5 5 5 5 5

20 20 20 20 20 20

0 1 2 3 4 5

1 2 3 4 5

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

93

Table 4 Parameters for thermal conductivity K, and total heat transfer rate Q. Q

=

Heat flux or total rate of heat transfer (W)

A k x T1–T2 m CP T6–T5

= = = = = = =

Cross sectional area in mm2 Effective thermal conductivity in W/m-K Sample thickness in mm2 Temperature difference between surfaces in
C Mass flow rate of cooling water used given as 0.01 Kg/s Specific heat capacity of liquid water given as 4.2 KJ/Kg/K Temperature difference in
C, between outlet and inlet flow of the cooling water measured using glass thermometers

2.4. Sample characterization

2.4.3. Morphological study

2.4.1. Thermal conductivity Steady state method using cussons thermal conductivity analyzer was applied in measuring thermal conductivity of the composites and composites according to ASTME 1530 standard. Square shaped samples of dimensions 10 mm  10 mm  3 mm were held under uniform compressive load between two polished surfaces, each controlled at a different temperature. For one dimensional heat flow, the difference in temperature and thickness of samples were used in calculating the thermal conductivity using Eq. (2).

2.4.3.1. Field emission scanning electron microscope (FESEM). The dispersion of GNP was observed through field emission scanning electron microscopy (FESEM) and transmission electron microscopy (TEM). FESEM micrographs of the fractured surface of uncompatibilized PP/GNP and novel PP/KF/MAPP/GNP hybrid composites were obtained using HITACHI-SU8020 and ZEISS SUPRA35VP. The uncompatibilized PP/GNP and compatibilized PP/ KF/MAPP and novel hybrid PP/KF/MAPP/GNP composites were gold coated using Balzers union MED 010 coater and carbon conductive adhesive 502. FESEM micrographs were effectively collected.

Q¼

kAðT1  T2Þ x

ð2Þ

The conduction heat transfer rate or heat flux (Q) is calculated by measuring the mass flow rate and temperature rise of the cooling water supplied at the lower end of the bar. At steady equilibrium state conditions, the total heat transfer rate Q, is given by Eq. (3) and parameters elucidated in Table 4.

2.4.3.2. Transmission electron microscopy (TEM). TEM micrographs were attained with samples of dimension 0.2 cm  1 cm via ZEISS LIBRA 120 Transmission Electron Microscope. The ZEISS LIBRA 120 transmission electron microscope (TEM) is equipped with an energy-filter and a Gatan-Ultrascan 1000 2 k  2 k CCD-camera operating at 120 KV accelerating voltage. 2.5. Thermogravimetry analysis

Q = mCp (T6–T5) (3) 2.4.2. Differential scanning calorimeter The melting and crystallization behavior of the composites were characterized using Mettler Toledo differential scanning calorimeter (DSC), with 5–10 mg samples sealed in aluminium crucibles. The temperature was raised from 30 to 300
C at a heating rate of 10
C/min. After a period of 1 min, temperature was swept back at rate of 10
C/min. The fusion enthalpy (DHf) of (PP) was measured and the degree of crystallinity Xc (PP) was calculated using Eq. (4). %X c ðPPÞ ¼

DHf ðPPÞ 1  100  DH0f ðPPÞ W ðPPÞ

ð4Þ

where, DH0f (PP) = 209 J/g (fusion enthalpy of 100% crystalline PP), W(PP) = combined weight fractions of PP + MA-g-PP

Thermogravimetry analysis (TGA) was carried out using NETZSCH TG 209 F3 TGA with samples of weight 5–10 mg for each run. The thermal decomposition of each sample was investigated under nitrogen atmosphere. Each sample was heated to 600
C at ambient temperature using heating rate of 10
C/min and weight loss recorded as function of temperature. 2.6. Heat deflection temperature (HDT) measurement Heat deflection temperature is defined as the temperature at which standard test samples deflects a specific distance under an applied load using ASTM D648. During testing, the bar samples of dimensions 120 mm  4 mm  4 mm were placed under GOTECH GT-HV2000 W HDT/VICAT measurement apparatus. A load of 0.46 MPa (66 psi) was initially applied, followed subsequently by a higher load of 1.80 MPa (264 psi). The specimens were subsequently

Fig. 1. FESEM micrographs of PP/KF/MAPP/GNP 1.

94

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

Fig. 2. FESEM micrographs of PP/KF/MAPP/GNP 3.

lowered into a bath containing silicon oil with intermittent temperature raise at 2
C per minute until deflection occurred at 0.010 inch (0.25 mm). Data of temperature at each specified load and deflection was collected. 2.7. Mechanical properties The tensile bar samples were tested using the universal materials tensile tester EZ 20KN, LLYORD instrument according to ASTM D 638 standard at a cross head speed of 50 mm/min. Five samples from each composition were tested along with neat blend specimens for comparison. The three-point loading applying centre loading technique was applied for flexural test (ASTM D790). Samples for flexural tests were placed on a supported beam. The span distance was set at 50 mm and the strain rate (compressed speed) was 3 mm/min. Impact test (ASTM D256) was performed using Izod Toyoseiki (11 J) impact tester at ambient temperature. 2.8. X-Ray diffraction analysis X’Pert, X-ray diffractometer (Siemens XRD D5000) and Nifiltered Cu Ka radiation at an angular incidence of 10–80 (2u angle range) was used in collecting X-ray diffraction patterns. XRD scans of GNP powder was collected at 40 kV and 50 mA with an exposure time of 120 s. Diffraction patterns were obtained using Bragg’s Eq. (5) to evaluate the dispersion of GNP and KF in matrix. nl = 2D sin up (5) where up = diffraction angleof the primary diffraction peak, D = distance between wavelengths, and l = X-ray wave length = 0.154 nm (Cu Ka), n = 1.

2.9. Fourier Transforms Infrared Spectroscopy Fourier transform infrared spectroscopy (FTIR) was conducted using a PerkinElmer 1600 infrared spectroscopy in studying the chemical interaction between the GNP and matrix by the KBr method using ratio of 1:100 prior to thin pellet conversion. FT-IR spectra of the coated pellet were recorded using a Nicolet AVATAR 360 at 32 scans with a resolution of 4 cm1 and within the wave number range of 4000 cm1–800 cm1. The points of notable transmittance peaks were determined using the “find peak tool” provided by Nicolet OMNIC 5.01 software. 2.10. Water absorption and dimensional Stability The water absorption and thickness swelling were conducted according to ASTMD 570. Prior to test, the dimensions of each specimen relative to length, width, thickness of each specimen were measured. Pre-conditioned samples were soaked in distilled water for 24 h at ambient temperature and at 50
C in a hot bath. Sample measurements were re-conducted after drying. Values reported for each sample are mean of five samples. 2.11. Thermal mechanical analysis (TMA) The coefficient of thermal expansion (CTE) quantitatively evaluates the expansion of a material over a temperature interval. It is very essential during manufacture of composites to ensure that composite components exhibit similar CTE values in order to eliminate building-up of thermal stresses, dimensional irregularities, prevent leakages and malfunctioning of composite components. The thermal expansion and CTE tests of the bionanocomposites were conducted via a thermomechanical analyzer (TMA 2940-TA Instruments) commencing from ambient

Fig. 3. FESEM micrographs of PP/KF/MAPP/GNP 5.

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

temperature to 250
C at a heating rate of 2
C min1 in a nitrogen atmosphere. Expansion mode of constant compressing load of 0.05 N was applied to the specimen in the testing process. Square shaped specimens of dimension 10  10  3 mm3 were used in conducting the tests. 2.12. Density measurements ASTMD 792 was applied in measuring density of PP/KF/MAPP/ GNP 1–5 phr nanocomposites by displacement 3. Results and discussion 3.1. FESEM analysis Figs. 1–3 shows FESEM micrographs of PP/KF/MAPP/GNP 1, PP/ KF/MAPP/GNP- 3 and PP/KF/MAPP/GNP-5 phr nanocomposites. In Fig. 1a and b, we see how effective compatibilization could result in better state of dispersion with the absence of agglomerates. We see also in Fig. 2(a, b) the micrographs of 3-phr fractured surface of the nanocomposites. We can observe the platelets embedded intact and dispersed within the matrix without signs of agglomeration at 3-phr nanocomposites. Sheets of GNP can be observed protruding out of the impact fractured surface. Apparently, they appear comingled with the matrix. Similar observation were reported in recent studies [19,20]. The homogenous dispersion of GNP in the matrix at 3 phr is believed to be responsible for the enhancement of tensile and impact strengths as elucidated in mechanical properties discussion later. This is unlike what we observe in Fig. 3 for the 5 phr concentration characterised by agglomerates and fibre pull-outs. The agglomeration of GNP, fibre-pull outs and voids observed in the matrix at this filler concentration is thought to be responsible for the decline in tensile and impact properties at this level of filler concentration as shown in later discussion. 3.2. Transmission electron microscopy (TEM) The level of distribution of nanoparticles in polymer is directly related to degree of improvement of inherent properties of nanocomposites such as thermal, mechanical, heat deflection and other properties. TEM images of nanocomposites were

95

collected from 0.2 cm  1 cm thin sections of 3 phr and 5 phr GNP specimens for better insight into degree of dispersion of nanoplatelets in the matrix. Fig. 4(a), and (b) depicts micrographs of thin sections of samples at 3 phr and 5 phr respectively. Apparently, as we can observe in Fig. 4(a), the presence of multilayers of graphene sheets is established, forming a continuous inter-connected network in presence of kenaf fibers as depicted by arrows. This shows evidence of exfoliation of graphene nanoplatelets and homogenous dispersion of GNPs in the matrix at PP/ KF/MAPP/GNP-3 phr. Fig. 4(b) shows agglomeration of graphene nanoplatelets as a result of filler–filler interaction and strong shearing GNPs were subjected to during melt extrusion processing. This observation is similar with previous work by Inuwa et al. [31]. Though, the GNPs were homogenously distributed in the polymer, full exfoliation has also been substantially attained for this system using the direct melt-processing technique. It is remarkable to note that inclusion of Ma-g-PP has promoted effective compatibilization and uniform dispersion of graphene nanoplatelets in PP/KF/MAPP/GNP-3 phr nanocomposites. In Fig. 4(a) GNP sheets were observed to exfoliate and uniformly disperse in polymer matrix with intercalation of fibres as depicted in fractured surface of TEM images. Some researchers reported that Ma-g-PP promoted filler dispersion in matrices through improved adhesion which enhanced wear and toughness of matrix on interaction [6]. 3.3. Heat deflection temperature and density Heat deflection temperature (HDT) of a polymeric material is an evaluation of its heat resistance towards applied load. Fig. 5(a) shows the dependence of the HDT on the GNP weight fraction for PP/KF/MAPP/GNP 1–5 phr nanocomposites. Here, it is observed that increasing inclusion of GNP into PP/KF/MAPP resulted in almost a monotonic increase in HDT values of PP/KF/MAPP/GNP nanocomposites. It is thought that inclusion of GNP and kenaf fibres resulted in blocking of the macromolecular chain mobility of polypropylene as highlighted by a corresponding trend in DSC Tg results in later discussion. We can observe from Fig. 5 that increasing inclusion of GNP in composites efficiently increased heat deflection temperature at 0.46 MPa (264 psi) from 85
C of pristine PP copolymer to 138
C at 5 phr loading. At 3 phr, the

Fig. 4. TEM images of hybrid PP/KF/MAPP/GNP at (a) 3 phr and (b) 5 phr.

96

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

Fig. 5. (a) Heat deflection temperature (b) Density graphs of PP/KF/MAPP/GNP.

highest heat distortion temperature value was observed. A lower but similar trend was observed at 1.8 MPa (264 psi) loading. When compared with the 0.46 MPa loading, a decreasing value of HDT was observed at all levels of filler concentration. But HDT increased from 60
C for neat PP to 89
C at 5 phr. This trend portends that synergy of GNP/KF inclusion has improved heat distortion temperature of the polymer and PP/KF/MAPP/GNP nanocomposites due to enhancement of glass transition of the polymer. Density measurements in Fig. 5(b) revealed density values of biocomposites close to literature reported values for neat PP with increasing values with increasing inclusion of GNP.

3.4. Thermal resistance Thermal resistance is an important parameter for thermal interface materials. It determines the ability of the material to resist heat dissipation. Eq. (6) is applied in calculating thermal resistance of materials. K¼

x R

ð6Þ

where K is the thermal conductivity of sample (W/m-K), x is the sample thickness (m) and R is sample resistance between hot and cold surfaces (m2-K/W). Fig. 6(a) shows the graph for thermal resistance of the samples. The thermal resistance of PP reduced monotonically by 99% from 16.64 W/m2K to 0.198 W/m2K at 5 phr.

Fig. 6. (a) Thermal resistance and (b) Coefficient of thermal expansion of PP/KF/MAPP/GNP nanocomposites.

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

This is attributed to high thermal conductivity of graphene given as 5000 w/mK which enhanced heat dissipation of the materials. 3.5. Coefficient of thermal expansion The thermal mechanical analysis (TMA) technique of measuring coefficient of thermal expansion (CTE) is a vital tool for understanding the dimensional changes of materials, in addition to the thermal stresses caused by increasing temperature [21]. A decreasing CTE value of the nanocomposites implies that the materials undergo a lower dimensional change on varying climate conditions [21]. Fig. 6(b) shows the trend of CTE of PP/KF/MAPP/ GNP nanocomposites. We can observe that a synergy exist between kenaf fibre and GNP on the CTE of the composites suggesting that addition of GNP in the materials improved their dimensional stability. A decreasing CTE with increasing inclusion of GNP (1– 5 phr) at constant 20 wt.% kenaf fibre is observed in the composites. This trend is attributed to the low CTE (1 106
C1) and unique platelet morphology of GNP particles and kenaf fibers in the cross-section of injection moulded samples which propagated the formation of GNP alignment with increasing inclusion of GNP. Kalaitzidou et al. [4] had reported a 25% reduction for 3 vol.% GNP based polypropylene. The inclusion of kenaf fiber also inhibited the thermal expansion caused by consistent climatic changes. This implies that strong bonding between the hydrophilic kenaf and the hydrophobic PP induced by MAPP also facilitated the decrease of CTE values through restriction of polymer chain mobility. As we can observe from Fig. 6(b), the CTE value of pure PP, 197.7  106
C1, was decreased to 160  106
C1 on inclusion of 20 wt.% kenaf fiber in PP/KF/MAPP biocomposites, showing a decrease of 18% in CTE. On inclusion of 5 phr GNP, CTE of the nanocomposites reduced to 40  106
C1 displaying a 79% decrease. The highest decrease in CTE was exhibited by the 3 phr concentration which reduced CTE to 38  106
C1, portending an 80% reduction when compared to CTE of neat PP. Another factor contributing to this decreasing trend would be the nanosize of GNP nanoplatelets. CTE of composites decreases with filler size in the case of inorganic reinforcing phases [4].

97

3.6. Thermal conductivity Heat transport is a critical issue for thermal interface materials applied between heat sources and heat sinks, which are essential ingredients of thermal management. Fig. 7(a) shows the trend of thermal conductivity of PP/KF/MAPP/GNP nanocomposites. The thermal conductivity of neat PP increased monotonically by 88%, from 1.8024  101 W/m K to 15.1298  101 W/mK at 5 phr inclusion of GNP. This improvement in thermal conductivity is attributed to the high thermal conductivity of graphene at 5000 W/ m K which enhanced the materials heat dissipation properties. Fig. 7(b) shows the thermal conductivity enhancement (K  K0)/ K0 at steady state as a function of the filler loading, and the results confirm the improvement of the thermal performance of the composites with increasing inclusion of GNP. The highest value of (K  K0)/K0 is achieved for GNP-5 phr (>1400% at 5 phr loading, where K is 15.1298  101 W/m K) and corresponds to an enhancement of about 100% per 1 wt.% loading, an outstanding value for filler efficiency. The thermal conductivity enhancement increases monotonically with increasing inclusion of GNP. 3.7. Water absorption and dimensional stability Fig. 8(a) and (b) reveals results of water absorption, and thickness swelling of composites at ambient temperature, and at 50
C, with different percentages of GNPs loadings. As we can observe, neat PP does not absorb moisture due to its hydrophobic nature, indicating that moisture is absorbed by the hydrophilic kenaf component in the composite as well as voids and micro-gaps at the interface. The value of water uptake was suddenly increased after the addition of KF to the blend. As the content of KF was constant (20 wt.%) in all materials, the different water absorptions among all the manufactured composites can be attributed to the role of GNPs. The water uptake was decreased after addition of GNPs. PP/KF/MAPP/GNP-3 phr showed lowest water uptake with 0.16% followed by PP/KF/MAPP/GNP 1 with 0.18 wt.%. The highest water uptake in the composites is PP/KF/MAPP/GNP-5 phr. This may be attributed to fiber pull-outs, voids and holes caused by agglomeration at this filler concentration evidenced by morphological studies. The graphene layers provided tortuous path and increased the barrier property for water transport [20]. Initially the

Fig. 7. (a) Thermal conductivity and (b) Thermal conductivity enhancement.

98

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

Fig. 8. Water absorption and thickness swelling at (a) 25
C and (b) 50
C.

cell wall is saturated by water through porous tubular and lumens of the fiber [14]. This is subsequently followed by water occupation of void spaces. Since nanocomposite voids and the lumens of kenaf were filled with GNPs, the penetration of water into the inner parts of materials by capillary action was suppressed. Another factor militating against water absorption could be the hydrophobic and water repelling nature of GNPs surface which immobilized some of the moisture, inhibiting the water permeation into the polymer matrix. From the trend of thickness swelling of the composites after immersion in distilled water for 24 h, we can observe that PP/ KF/MAPP biocomposite without GNPs exhibited the highest thickness swelling values among other specimens. Similar to water uptake results, the composites with GNP inclusion exhibited less thickness swelling percentage in comparison with those fabricated without it. Thus, this phenomena and elucidation of thickness swelling of samples were similar to those of samples discussed for studies of water uptake. No significant difference was observed for water absorption and thickness swelling at ambient temperature and 50
C respectively. This is attributed to the role of GNP.

at 1774 cm1 and 1778 cm1 are assigned to the symmetric C¼O stretching of maleic anhydride (MA) functional groups on MA-g-PP [22] indicating that functional groups of MA reacted with the hydroxyl group of kenaf in PP/K/MAPP/GNP nanocomposites forming covalent bond and esterification reaction [23]. The peaks are characteristics of the diametric form of a dicarbonylic acid and cyclic anhydride [22]. This is confirmed by low stretching vibration of ester carbonyl groups (C¼O) between 1741 cm1 and 1739 cm1 which have resulted in the esterification reaction between free OH groups of kenaf fiber and the MA functional groups of MA-g-PP [27]. Thus, ester bonding in the hybrid nanocomposites provided better wettability and dispersion which enhanced mechanical and thermal properties. The band at 2360.87 represents the medium intensity of carboxylic O-H stretching vibration while the bands between 2916 cm1 and 2949.16 cm1 represent the stretching of the C H bonds. The absence of graphene and graphene oxide (GO) bands reveals the purity of GNP used in this study. These characteristic strong graphene and GO peaks are situated at 3400 cm1 (O H) stretching vibrations, 1720 cm1 representing C¼O stretching vibrations, 1220 cm1 representing the C-OH

3.8. FTIR analysis In order to observe any occurring chemical changes, the samples were analysed using FTIR spectroscopy. Fig. 9 displays the FTIR spectra of PP/GNP and PP/KF/MAPP/GNP nanocomposites. From Fig. 9 it can be observed that no visible peaks are observed in the spectrum of GNP. The FTIR spectra data of the sample reveals the presence of various linkages such as ester group, carboxylic acids, oleifinic double bonds and other characteristic peaks. The peaks at 800 cm1 are attributed to C H outer bending vibration, CH in-plane bending and out-of-plane wagging. [22], while peaks at 977.12–997.2 cm1 are attributed to the strong CH and O H bending of carboxylic acids [19]. The peak absorptions at 1375.2 cm1 are the medium C H rock vibrations of alkanes [23]. The characteristic absorption peaks at 1456 represent the C H bending vibrations of alkenes, while 1558 cm1 represent the strong intensity of C¼C stretch vibration of aldehydes functional groups. The biocomposite peak at 1560 cm1 represents CCC symmetric stretching while the peak at 1653 cm1, represents the keto form stretching frequency of C¼O. The low absorption peaks

Fig. 9. FTIR spectra of PP/GNP and PP/KF/MAPP/GNP nanocomposites.

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

stretching, and 1060 cm1 representing the C-O stretching [24]. This implies that the interactions occurring between GNP and kenaf fibers are mainly physical because GNP has nil polar groups. Similar observations have been reported [13–26]. Thus, no chemical interaction occurred between the polymer matrix and GNP due lack of any structural changes in the composites. Hence, any property improvements in the composites are ascribed to physical interactions between GNPs, and kenaf/MA-g-PP/polymer matrix system which enhanced interfacial adhesion of matrix to GNP sheets. Similar observations were also reported elsewhere [13–26].

99

3.9. X-ray diffraction analysis The XRD characteristic peaks of GNP and kenaf fiber are shown in Fig. 10(a) and (b) respectively. X-ray diffraction (XRD) facilitates detection of interlayer spacing between graphene sheets and exfoliated state of graphene. As we can observe in XRD diffractogram in Fig. 10(a) for pristine GNP, a strong sharp reflection at 2u = 26.5
, corresponding to interlayer spacing of 0.34 nm, and unique spacing of graphene units in 0 0 2 plane is seen clearly. The XRD patterns of GNP, uncompatibilized PP/GNP-5 phr and hybrid

Fig. 10. XRD crystallographic patterns of (a) GNP and (b) kenaf fiber.

100

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

Fig. 11. XRD patterns of GNP, PP/GNP-5phr and hybrid PP/KF/MAPP/GNP (PPKMG3phr and PPKMG-5phr) nanocomposites.

PP/KF/MAPP/GNP-3 phr, and 5 phr nanocomposites are shown in Fig. 11 . Fundamentally, thermoplastic polymers undergo crystallization to specific crystalline forms. For isotactic PP, there are various packing geometries of PP-helices resulting in four established crystal structures including a-monoclinic, b-trigonal, g-triclinic and d-smectic crystallographic forms depending on melting chronology, crystallization temperature, presence of extraneous materials and cooling rate [7] though the most common crystal is the a-monoclinic crystal form [27]. However, reports have revealed that the b-form exhibit superior impact strength and toughness due to its unique lamellar morphology [7] strain induced b-transition during mechanical deformation [28] and formation of large plastic region. The peak observed at 2u = 16.91
in the diffraction pattern of the nanocomposites, corresponding to the characteristic peak [3 0 0] plane of b-form crystals is due to inclusion of GNP in the hybrid PP/KF/MAPP/GNP nanocomposites. A similar observation was reported elsewhere [7]. The second b-phase peak observed at 2u = 21.88
is for [3 0 1] crystallographic plane [7]. Thus, it can be concluded that synergistic inclusion of both graphene nanoplatelets and kenaf fibers in PP effectively enhances nucleation of b-crystals which exhibit superior impact strength and toughness [7] and provide secondary mechanism of reinforcement. The inclusion of GNP induces formation of another peak at 2u = 26.37
corresponding to a [0 6 0] which also initiates formation of b-crystals at 2u = 16.91
, [7] though the peaks at 2u = 26.37
(graphene 0 0 2 plane) in GNP, PP/GNP-5 phr and PPKMG-5 phr respectively corresponding to interlayer spacing of 0.34 nm are obviously absent in the diffraction pattern of PP/KF/ MAPP/GNP-3 phr nanocomposites, implying a complete exfoliation of GNP at this filler loading. It is observed that presence of kenaf fibers and graphene sheets influences crystallization

behavior of matrix with some variations in the crystal face of {[11 0], [0 4 0], [1 3 0]} [7]. Thus, the strong nucleating effect of GNP can be observed through increase of peaks corresponding to planes matching the crystallographic plane of GNP. These peaks are situated at 2u = 26.37
corresponding to a [0 6 0] as previously stated. This can be attributed to the homogenous dispersion and partial exfoliation of GNP sheets in PP matrix and the nucleating influence of GNP which also increases crystallization temperature [7]. Additionally, inclusion of GNP influenced the diffraction pattern of the nanocomposites evidenced by the peaks becoming narrower and taller in PP/GNP-5 phr and PP/K/MAPP/GNP-5 phr nanocomposites compared to the broad peaks of PP/KF/MAPP/ GNP-3 phr nanocomposites. This indicates that crystals are becoming thinner and more homogeneous. This is probably due to a more homogeneous distribution [7] and exfoliation of GNP sheets in the matrix. Overall, as shown in Fig. 11, at low inclusion of GNP-3 phr in PP/Kenaf nanocomposites, disappearance of characteristic GNP peak implies complete exfoliation and good dispersion of graphene nanosheets in nanocomposites. For nanocomposites containing higher GNP concentration at 5 phr, appearance of strong graphene peaks indicate homogeneous distribution and partial exfoliation of GNP in PP/KF/MAPP/GNP nanocomposites. Similar observations are also reported [27]. 3.10. Differential scanning calorimeter The synergistic effect of GNP and KF on thermal behavior of PP, PP/KF/MAPP biocomposites, PP/GNP and PP/KF/MAPP/GNP nanocomposites were studied using differential scanning calorimetry (DSC). The thermal data is shown in Table 5. Fig. 12 shows heating and cooling scans of DSC traces for the materials. 3.11. Flexural strength and modulus properties The flexural behavior of PP/KF/MAPP/GNP nanocomposites are shown in Fig. 13. We can observe in Fig. 13(a) that flexural modulus exhibited steady increase with increasing GNP loading. This is similar to trends reported previously [1–5,13,20] and values for flexural modulus increased from 963 MPa for pristine PP and 1905 MPa for PP/KF/MAPP biocomposite to 33 GPa for 5 phr of PP/KF/MAPP/GNP nanocomposites. These shows 95% and 98% increases respectively. These increasing trends are attributed to the high stiffness, uniform and homogeneous dispersion of GNP in the polymer matrix as evidenced by morphological images in Figs. 1, 2 and 4(a) respectively. Additionally, improved bonding between polymer matrix, micro-natural fiber, and nano-fillers may also be attributed to efficient compatibilization provided by maleic anhydride grafted polypropylene (MA-g-PP) which resulted in restriction of matrix deformation at the elastic region which improved modulus. The high enhancement of flexural modulus may also be attributed to the alignment of graphene nanofillers to direction of flow during extrusion and injection molding cycles [13].

Table 5 DSC OF PP, PMK, PP/GNP-3 phr and PKMG 1–3 phr nanocomposites. Designation

Crystallization Temperature (
C) PP

Melting Temperature (
C) PP

Crystallinity (%) PP

PP P/KF/MAPP PP/GNP PP/KF/MAPP/GNP 1 PP/KF/MAPP/GNP 3 PP/KF/MAPP/GNP 5

100  0.5 101  0.5 105  0.3 106  0.1 110  1.0 111  1.02

159.80  1.2 160.96  0.13 160.46  63 160.66  0.43 161.54  0.44 163.55  2.0

21.98  0.5 20.69  0.2 20.47  0.3 18.94  0.3 16.87  0.1 15.96  0.2

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

101

morphological studies displayed in Figs. 3 and 4(b). Some previous studies reported similar trends [7–14]. 3.12. Young’s modulus and tensile strength

Fig. 12. Heating and Cooling scans of PP, PP/KF/MAPP biocomposites and PP/KF/ MAPP/GNP nano-biocomposites.

From Fig. 13(b) we can see that flexural strength displayed a moderate improvement up to 3 phr, after which a steady decrease was observed. The flexural strength improved from 14.2 MPa for neat PP and 26.7 MPa for PP/KF/MAPP to 48.2 MPa at 3 phr level of filler concentration, beyond which a decline was observed. This portends an improvement of 70% and 45% respectively compared to pristine PP. Additional incorporation of GNP resulted in a decline as observed at 46.7 MPa for 5 phr; some previous studies reported similar observations [13–15]. The flexural strength mechanism of decline beyond 3 phr may be ascribed to escalating incorporation of GNP which minimized melt flow of the polymer and inhibited plastic flow which hindered the polymer free flow thereby inhibiting mobility of polymer chain segment, as a result of agglomeration which induced increased shear stress and large specific area of GNP in the matrix [7]. These phenomena may have escalated filler–filler interaction, and nanofillers stacking, resulting in uneven dispersion of nanofillers, and non-uniform distribution of fibers, voids and fiber-pull-outs in the polymer matrix thereby minimizing flexural strength as supported by

The graphs for Young’s modulus and tensile strength of nanocomposites are shown in Fig. 14(a) and (b) respectively. We can observe that Young’s modulus steadily improved by 56% at 5 phr inclusion of GNP, from neat PP value of 706 MPa to 1600 MPa. PP/KF/MAP/GNP nanocomposites displayed enhanced Young’s modulus due to optimum interfacial strength attained through inclusion of higher content maleic anhydride grafted polypropylene (MA-g-PP) derived through exfoliation mechanism which also restrained the matrix from deformation in the elastic region. This phenomenon improved modulus by enhancing uniform distribution of GNP within the polymer phase. This is evidenced by morphological images shown in Figs. 1, 3 and 4(b) respectively. Also from Fig. 14(b) tensile strength improved moderately to 21 MPa at 3 phr for PP/KF/MAPP/GNP nanocomposites from 15 MPa for neat PP showing an increase of 30%. Tensile strength declined beyond 3 phr because of filler–filler interaction due to agglomeration, and other defects such as holes, and fiber pull-outs as shown by morphological images in Figs. 3 and 4(b). 3.13. Izod impact strength The graph for notched Izod impact strength is displayed in Fig. 15(a) and shows the effect of increasing inclusion of GNP on notched impact strength of hybrid PP/KF/MAPP/GNP nanocomposites. From the graph we can observe a minor decline of impact strength of neat PP in comparison to control PP/KF/MAPPbiocomposite from 5628 J/m2 to 5456 J/m2. But on low inclusion of GNP in PP/KF/MAPP/GNP nanocomposites, a 12% improvement in impact strength was observed up to 3 phr at 6218 J/m2. A recent study reported a similar trend [7]. This is possible because GNP can initiate nucleation of b-PP crystals. The rapid decline in impact strength observed beyond 3 phr filler concentration is attributed to the incompatibility between matrix and filler, in addition to the heterogeneity of PP/KF/MAPP/GNP nanocomposites. Also, higher inclusion of GNP led to increased density of nanocomposites

Fig. 13. Effect of GNP loading on (a) flexural modulus and (b) flexural strength of PP/KF/MAPP/GNP nanocomposites.

102

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

Fig. 14. Effect of GNP on (a) Young’s modulus and (b) Tensile strength of PP/KF/MAPP/GNP nanocomposites.

resulting in a fragile brittle material with low toughness which caused decrease of impact strength. A reduction in elongation was observed as shown in Fig. 15(b). 3.14. Thermal stability Thermal stability of pristine polymer, biocomposites and nanocomposites were studied under nitrogen gas atmosphere by thermogravimetry. The TGA and DTG curves of specific samples are shown in Figs. 16 and 17 respectively, and values reported in Table 5. The results show the effect of increasing inclusion of graphene nanoplatelets on the thermal stability of PP/KF/MAPP/GNP hybrid nanocomposites. From Table 5,T20, T50, and Tmax are notably high for all the composites containing GNP compared to pristine PP and control PP/KF/MAPP biocomposites. Interestingly, the 3 phr

composite exhibited highest thermal stability. This is attributed to high aspect ratio of GNP which serve as barrier restricting emission of gaseous molecules during thermal decomposition. The improvement at 3 phr is because of the homogenous and uniform distribution of GNP sheets at this filler concentration. Uniformly dispersed GNP repressed the supply of oxygen by forming layers of char on the surface of composites thereby enhancing thermal stability [7]. The highest amount of char residue was also observed at this threshold. This means superior enhancement of thermal stability of nanocomposites at this filler level (Table 6). 4. Conclusion The effects of the synergy between hybrid GNP/kenaf systems on material properties of PP/KF/MAPP/GNP nanocomposites were studied. PP/KF/MAPP/GNP hybrid nanocomposites were

Fig. 15. (a) Effect of GNP on notched Izod impact strength of PP/KF/MAPP/GNP nanocomposites (b) Elongation at break.

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104

Fig. 17. DTG curves of PP, PP/KF/MAPP biocomposites, PP/GNP and PP/KF/MAPP/ GNP nanocomposites.

103

Fig. 16. TGA curves of PP, PP/KF/MAPP biocomposites, PP/GNP and PP/KF/MAPP/ GNP nanocomposites.

Table 6 TGA data of PP, PMK, PP/GNP 3-phr, PKMG 3-phr nanocomposites. C

Tonset (
C)

T20 (
C)

T50 (
C)

DTG Peak Temp Tmax (
C)

Tend (
C)

Residual weight (%) at 600
C

PP PMK PP/GNP -3 phr PKMG-3 phr

442  0.72 448  0.12 452  0.27 455  0.57

450  1.45 463  0.15 465  0.05 480  1.55

460  1.1 465  0.6 470  0.12 490  1.8

473  0.5 474  0.4 475  0.3 490  1.2

488  0.25 489  0.15 490  0.5 495  0.4

0.4  0.46 0.47  0.45 1.18  0.38 18  1.29

successfully fabricated via eco-benign melt processing technique using co-rotating twin screw extruder. Morphological and structural studies using FESEM show uniform distribution of GNP in the matrix. Maximum flexural, tensile and impact strength were obtained at GNP loading of 3 phr. Enhancement of mechanical properties were attributed to high graphene platelet modulus (1TPa), efficient stress transfer between matrix, microfibrous fillers and GNP, in addition to homogeneous and uniform dispersion of GNP in matrix. XRD analysis reveals that the presence of both graphene nanoplatelets and kenaf fibers enhanced nucleation of b-PP crystals which inculcate superior impact strength and toughness and provide secondary reinforcement mechanism. At low inclusion of GNP in nanocomposites, appearance of characteristic GNP peak implies complete exfoliation and good dispersion of graphene nanosheets in nanocomposites. At higher inclusion of GNP in nanocomposites, emergence of strong graphene peak indicates dispersion and partial exfoliation of GNP in polymer. FTIR analysis revealed that the interactions between GNP and PP/KF/MAPP system were mainly physical, and not chemical, because GNP has no polar groups. Hence, no structural changes were observed in the composites but enhanced interfacial interaction of matrix/kenaf to GNP sheets. Results from TGA show that graphene nanoplatelets significantly improved thermal stability of PP/KF/MAPP/GNP nanocomposites at 3 phr loading. Results from DSC indicated that GNP, KF and MAPP were strong nucleating agents for PP with almost no variation in melting temperature (Tm), increasing crystallization temperature (Tc), and decreasing degree of crystallization (%). The CTE reduced with increasing inclusion of nanofillers while HDT increased with inclusion of GNP. Thermal conductivity monotonically increased with increasing inclusion of nanoparticles. The propensity of materials to absorp water and dimensional instability were greatly reduced with increasing inclusion of GNP. Overall, parameters

influencing effective improvement of properties are uniform and homogeneous distribution of GNP layered nanoplatelets in the hybrid kenaf-polymer system. Drawbacks of the system include filler–filler agglomeration, cracking, wrinkling and overlapping of GNP platelets observed at high loading of GNP. The novel hybrid compatibilized GNP filled PP/KF nanocomposites have prospects of application in aerospace, automotive, construction and electronics industries where heat dissipation, strength, stiffness and toughness are essential. Acknowledgement The authors wish to acknowledge the management of Universiti Teknologi Malaysia for providing enabling infrastructures for the success of this research. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

X.G. Liu, Carbon 51 (2013) 438. A.K. Geim, K.S. Novoselov, Nat. Mater. 6 (2007) 183. C. Lee, W. Xiaoding, J.W. Kysar, J. Hone, Science 321 (2008) 385. K. Kalaitzidou, H. Fukushima, L.T. Drzal, Carbon 45 (2007) 1446. R.J. Young, I.A. Kinloch, L. Gong, K.S. Novoselov, Compos. Sci. Technol. 72 (2012) 1459. K. Kalaitzidou, H. Fukushima, L.T. Drzal, Compos. Part A-Appl. Sci. Manuf. 38 (2007) 1675. K. Kalaitzidou, H. Fukushima, H. Miyagawa, L.T. Drzal, Polym. Eng. Sci. 47 (2007) 1796. K. Tapas, B. Sambhu, Y. Dahu, H. Nam, B. Saswata, H. Joong, Prog. Polym. Sci. 3 (2010) 1350. C.I. Idumah, A. Hassan, A.C. Affam, Rev. Chem. Eng. 31 (2015) 149. Z. Xu, C. Gao, Macromolecules 43 (2010) 6716. S. Kim, I. Do, L.T. Drzal, Polym. Compos. 31 (2010) 755. A. Duguay, J. Nader, A. Kizilitas, J. Gardner, H. Dagher, Appl. Nanosci. 4 (2014) 279. S.O. Han, M. Karevan, M. Bhuiyan, H. Park, K. Kalaitzidou, J. Mater. Sci 47 (2012) 3535. I. Ghasemi, B. Kord, Iranian Polym. J. 18 (2009) 683.

104 [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

C.I. Idumah, A. Hassan / Synthetic Metals 212 (2016) 91–104 M. Biswal, S. Mohanty, S.K. Nayak, J. Appl. Polym. Sci. 125 (2012) 432. C. Majid, H. Yahya, E.et al Ghanbar, Polym.r Bull. 71 (2014) 349. X. Jiang, L.T. Drzal, Polym. Compos. 31 (2010) 1091. S. Lee, I. Kang, G. Doh, W. Kim, J. Kim, H. Yoon, Q. Wu, eXPRESS Polym. Lett. 2 (2008) 78. S. Sheshmani, A. Ashori, M. Fashapoyeh, Int. J. Biol. Macromol. 58 (2013) 1. Y.M. Chang, H. Kim, J.H. Lee, Y.M. Song, Appl. Phys. Lett. 97 (2010) 211102. B. Debelak, K. Lafdi, Carbon 45 (2007) 1727. M. Sclavons, M. Laurent, J. Devaux, V. Carlier, Polymer 46 (2005) 8062. F. Marti-Ferrer, F. Vilaplana, A. Ribes-Greus, A. Benedito-Bornas, C. Sanz-Box, J. Appl. Polym. Sci. 99 (2005) 1823. Z. Xu, C. Gao, Macromol 43 (2010) 6716.

[25] C.M. Chan, J. Wu, J.X. Li, Y.K. Cheung, Polymer 43 (2002) 2981. [26] X. Wen, Y. Wang, J. Gong, J. Liu, N. Tian, Y. Wang, Z. Jiang, J. Qiu, T. Tang, Polym. Degrad. Stab. 97 (2012) 793. [27] J.H. Chen, F.C. Tsai, Y.H. Nien, P.H. Yeh, Polymer 46 (2005) 56808. [28] J. Karger-Kocsis, J. Polym. Eng. Sci. 36 (1996) 203. [29] A. Najafi, B. Kord, A. Adi, S. Ranaee, J. Thermoplast. Compos. Mater. 25 (2012) 717. [30] H. Rozman, L. Musa, A. Azinwati, A. Rozyanty, J. Appl. Polym. Sci. 119 (2010) 2549. [31] M. Inuwa, A. Hassan, D. Wang, Samsudin, M. Mohamad Haafiz, S. Wong, M. Jawaid, Polym. Degrad. Stab. 110 (2014) 137–148.