Mechanical and thermal behaviours of graphite flake-reinforced acrylonitrile–butadiene–styrene composites and their correlation with entanglement density, adhesion, reinforcement and C factor

Abstract

Polymer composites are prepared by melt compounding of acrylonitrile–butadiene–styrene (ABS) and graphite flakes (GFs). The fabrication of ABS/GF composites involves two steps: melt compounding of the ABS matrix and GFs in a vertical twin screw micro-compounder for a period of 3 minutes, and then sample preparation using a micro-injection moulding machine. The GF content is varied from 0 to 40 vol% with respect to the ABS matrix. The composites are characterized through tensile, flexural, impact, hardness and thermal conductivity. The thermo-mechanical behaviour is analysed by dynamic mechanical thermal analysis (DMTA). The ABS/GF composites are found to have a gradual variation in the flexural modulus, with about 6, 25 and 92% increments in the flexural modulus on addition of 1, 9 and 40 vol% of GFs in the polymer matrix, respectively. But this addition of GFs exhibits a small decrement in the tensile strength and elongation at break. The impact strength is also sharply decreased (40% reduction with only 1% GF) with the increasing GF content in the ABS matrix. The DMTA results show an improvement in the storage modulus of ∼ 250% at room temperature, and the loss modulus also increases while the damping factor decreases as the GF content increases in the composite. The degree of entanglement increases whereas the reinforcement efficiency and C factor decrease for higher GF content. These are calculated from the data obtained by DMTA. The thermal conductivity of the composites shows an increasing behaviour with an increasing amount of GFs, as a ∼250% increment is observed at the 40 vol% loading of GFs. The addition of GFs also increases the melt viscosity, and the same trend is shown by the complex viscosity. Thermogravimetric analysis shows an improvement in the thermal decomposition temperature, and the char yield shows a 35% improvement at 40 vol% loading of GFs. Therefore the GF-reinforced ABS composite with improved thermal conductivity, heat stability, viscoelastic behaviour and flexural modulus can be a promising as well as a suitable composite material for making various electronic and electrical accessories including bipolar plates for fuel cell applications.

---

Introduction

Improvement in the performance of a polymeric component/product can be achieved by fabricating polymer-based composite materials. A polymer composite consists of a polymer as a matrix and filler as reinforcement or functional material. These polymer composites have found applications in areas such as aerospace, automotive, chemical industries, electronic enclosures, etc.^1–3 A number of modifications can be done in the polymer matrix to improve the performance, from inclusion of micro to nano-fillers, and blending of two or more polymers with each polymer with or without filler addition. For preparation of polymer composites the choice of the fabrication process is as important as the selection of a proper composition of the constituent involved.^1,3 Recently Sudeepan et al. studied the effect of titania (TiO₂) in acrylonitrile–butadiene–styrene (ABS) composites and concluded that the mechanical and tribological properties improve up to 10 wt% of filler addition.⁴ Zinc oxide (ZnO), having prominent chemical and physical properties, has been used as an inorganic filler to improve the mechanical properties.⁵ Addition of ZnO in ABS polymers increases the tensile strength and flexural strength up to 15 wt% then starts decreasing, however the hardness keeps on increasing as observed by Sudeepan et al.⁵ Calcium carbonate (CaCO₃) has also been added in ABS.^6–8 The addition of CaCO₃ in the ABS matrix improves the stiffness and bending strength as well as the heat dissipation properties.⁶ A hollow glass bead-filled ABS polymer exhibits a linear increase in Young’s modulus. As the filler content increases beyond 5 vol%, the tensile as well as flexural strength start decreasing.⁹ Another study done by Kar et al. with addition of alumina (Al₂O₃) in ABS shows a slight increase in Young’s modulus and the flexural modulus, but a lower tensile and impact strength, flexural modulus and strain at break.¹⁰ However, Rai et al. utilized waste silk fabric to improve the mechanical properties of an epoxy–ABS blend.¹¹ The lightweight thermoplastic composites of polycarbonate/ABS copolymer alloys and chemically-treated recycled carbon fibre prepared by Yan et al. show significant improvement in the mechanical properties but the electrical resistance is reduced to some extent.¹²

Carbon black (CB) in the nano-form is used as reinforcement in rubber/elastomeric products and other polymers, as an electrically conducting filler in plastics, rubbers and batteries, and as black pigment or a UV-stabilizing agent in surface coatings, inks and plastics.^1,2 In the study of morphology, the mechanical and thermal behaviour of ABS/CB composites by Shenavar and Abbasi, it is revealed that CB particles are distributed in the polybutadiene phase and styrene–acrylonitrile phase of ABS, and the cross-linking in the polybutadiene phase during the processing controls the degradation process of the polymer matrix.¹³ Upon increasing the carbon black content in the composite, the thermal stability is increased while the impact strength is decreased.¹³ Electrically conducting nano CB composites with ABS–polycaprolactam blends studied by Mondal et al. have shown significant improvement in the electrical conductivity and mechanical properties depending on the loading level of the filler.¹⁴

Carbon nanotubes (CNTs) are also used as a filler material for the enhancement of properties in the polymer composites. The advantages of using CNTs as a filler are an improvement in the interfacial adhesion between the filler and matrix resulting in better composite properties.³ Mari and Schaller have studied ABS/CNT polymer composites and suggested that the addition of CNTs shifts the glass transition temperature (T_g) towards a higher temperature region and delays the modulus drop.¹⁵ Shrivastava et al. studied the percolation threshold of multi-walled carbon nanotubes (MWCNTs) in the ABS matrix.¹⁶ Localisation of MWCNTs in the polycarbonate phase has been seen by Sun et al. at a very low ABS content due to the higher affinity of MWCNTs.¹⁷ For a lower loading of MWCNTs a well dispersed nanocomposite using ABS prepared by Singh et al. via melt mixing shows improvement in the electrical conductivity, storage modulus and thermal stability.¹⁸ De Oliveira et al. have come up with a conclusion that the incorporation of up to 2 wt% MWCNTs in ABS is responsible for an improvement in the electrical conductivity and corrosion resistance, but it loses the thermal stability of the composite.¹⁹ Blending of polymers with CNTs has also shown improvement in the properties.^20–22 Wang et al. prepared ABS/MWCNT composites by solvent blending in cyclohexanone solvent and concluded that at a low loading of MWCNTs in the ABS matrix, the percolation point appears and the ABS matrix is destabilized by MWCNTs in the temperature range of 220–450 °C.²³ The effect of CNTs on the viscoelastic nature has been explored by Jyoti et al.²⁴ Jyotishkumar et al. justified the fact that for upgrading the properties of epoxy-based polymer composites, addition of thermoplastics is an effective tool.²⁵

On the other hand, graphite, the thermodynamically most stable form of carbon, is highly anisotropic in nature and is a good conductor of heat and electricity.² Difallah et al. studied the effects of graphite in ABS on the tribological and mechanical properties. On the addition of 7.5 wt% (8 vol%) graphite in the ABS matrix, the elastic modulus and failure strain decrease, whereas the wear resistance improves.²⁶ A comparative study of nanocomposites of styrene and its copolymers, high impact polystyrene (HIPS) and polystyrene (PS) using graphite nanopowder as a filler by Uhl et al. showed no improvement in the mechanical properties of ABS and PS-based composites, whereas there was some improvement in Young’s modulus.²⁷ Dahiya et al. confirmed the improvement in the electrical conductivity and dielectric properties of the ABS/graphite composite for 7.6 vol% graphite addition.²⁸ Wang et al. revealed that the addition of 5 phr exfoliated graphite into the ABS matrix is enough to decrease the coefficient of friction, with a slight increase in the tensile strength and hardness.²⁹ Sharma et al. observed the synergistic effect of graphene/MWCNTs in the ABS composite.³⁰ Apart from the filler materials and matrix, the prepreg geometry also governs the properties of the composite as shown by Mariatti et al.³¹ The aesthetic appearance of the final product made of ABS can be improved by using gas-assisted injection moulding.³²

ABS is a widely used engineering thermoplastic polymer and has characteristics like excellent surface finish, high impact resistance, high dimensional stability at elevated temperature, moderate tensile strength, good stiffness, impact strength, abrasion resistance and chemical resistance, but poor thermal stability, and thermal conductivity processability.^1,10,33,34 In order to overcome the poor thermal, electrical and mechanical performance, researchers have incorporated various inorganic materials such as TiO₂,⁴ ZnO,⁵ CaCO₃,^6–8 hollow glass beads,⁹ Al₂O₃,¹⁰ iron powder,³⁵ wood dust,³⁶ waste silk fabric,¹¹ carbon fibres,^12,37 carbon black,^13,14 CNTs,^{15,17–19,23,25} graphite,^26,28,29 textile fibre,³⁴ used tyres,³⁸ fly ash,³⁹ graphene,⁴⁰ etc. Out of them only a few researchers have added graphite in a small percentage (0–8 vol%) into the ABS matrix and reported some improvement in terms of the electrical conductivity and thermal stability but reduction in some mechanical properties. GFs are thermodynamically a very stable material having a very high thermal conductivity and high thermal stability, but are opaque and possess a low tensile strength, are brittle in nature and are very cheap compared to CNTs and graphene. In this context, GFs may be considered as a suitable ideal material for the ABS matrix to enhance the electrical and thermal conductivity and also the dimensional stability of the ABS matrix, which may be used for the manufacture of various electronic and electrical enclosures and other accessories. At the present time a lot of electronic items and electrical accessories are being made with the ABS polymer. As the mechanical and thermal properties are of much importance where electronic enclosures are concerned, the lower thermal conductivity may result in heating up of the devices and hence a lowering of their efficiency. The present work focuses on the feasibility of a GF-reinforced ABS polymer with improved thermal conductivity, thermal stability, mechanical and viscoelastic properties, which may be suitable for electrical devices such as electronic and electrical enclosures including bipolar plates for fuel cell applications⁴¹ and to establish a co-relationship between the properties with other controlling parameters like the entanglement density, adhesion, reinforcement and C factor.

Experimental

Materials used

The matrix material used in this work was ABS (Grade-MIF-45) procured from M/S Bhansali Polymers, Vadodara, Gujrat, India. The filler material GFs with a particle size of +100 mesh size (∼150 micron) was purchased from M/S Sigma-Aldrich, India.

Composite preparation

ABS/GF composites were prepared using the melt mixing technique. The mixing was performed using a vertical twin screw DSM Xplore® micro-compounder having a volumetric capacity of 5 millilitres. The presence of moister could affect the mixing as well as the properties of the prepared composites. Prior to the composite preparation the ABS polymer was kept in an air-circulating dryer at 80 °C for 8 h and the GFs were also dried for 8 h at 90 °C in an oven. The content of GFs was varied in the composite as 0, 1, 3, 6, 9, 12, 20, 30 and 40 vol% with respect to the matrix material. The processing of the composite was done in speed controlled mode at 230 °C with a screw speed of 120 rpm. After feeding, the material was allowed to process for 3 minutes at the same temperature. After processing, the melt mix was taken out in the form of strands and chopped into small pellets. These pellets were used to prepare five samples of each composition as per ASTM standards by a micro-injection moulding machine from DSM Xplore®. The barrel temperature and mould temperatures were set to 240 and 80 °C, respectively. The injection pressure, holding pressure and releasing pressure were set to 12, 10 and 10 bar, respectively while the injection time, holding time and ejection time were kept at 5, 13 and 2 s, respectively.

Characterizations

Dynamic mechanical thermal analysis (DMTA). To study the viscoelastic behaviour of the ABS/GF composites, DMTA (PerkinElmer, Pyris Diamond DMA (e7)) was used. The sample dimensions used were 40 mm × 10 mm × 1.6 mm. The test was performed in tension mode with a frequency of 1 Hz, strain amplitude of 5 μm under nitrogen atmosphere (flow rate of 200 ml min⁻¹) and temperature of −100 to 160 °C. The inbuilt software in DMTA was used to collect the complex modulus, storage modulus, loss modulus, damping, complex viscosity, storage viscosity and loss viscosity.

Melt viscosity. The inbuilt software of the DSM Xplore® micro-compounder was used to collect the data of melt viscosity while processing the polymer and GFs in the micro-compounder. On completion of feeding, the data for the melt viscosity were collected for 3 minutes.

Morphology. The morphology of the composite fractured surface via tensile test was imaged using FEI QUNATA 200 scanning electron microscopy (SEM). Samples were sputter-coated with Au prior to the SEM study.

Mechanical properties. Tensile samples were prepared according to ASTM D638. The test was performed at room temperature (23 ± 2 °C) using a universal tensile testing machine (UTM) Zwick Roell Z010 having a maximum load capacity of 10 kN. The gauge length for the test was fixed at 60 mm, while the cross-head speed used was 60 mm min⁻¹. For the flexural modulus, the samples were prepared as per ASTM D790. The test was performed at room temperature (23 ± 2 °C) using the same UTM as in the case of the tensile test. The cross-head speed was set to 50 mm min⁻¹ while the span length was 45 mm. The Izod impact test was conducted according to ISO 179 using an impact tester (model-Impact-104) from TINIUS OSLEN USA. The V-notch on the test samples were cut by a notch cutter from Ceast Italy.

Thermal conductivity. The measurement was carried out using the laser flash method using LFA 1000, Linseis (Germany). The test sample has a diameter of 25 mm and a thickness of 1.6 mm. The test was performed in the temperature range of 20 to 60 °C.

Thermogravimetric analysis (TGA). The analysis was performed in a Perkin Elmer TGA (USA) Instrument in the temperature range of 30 to 600 °C at a constant heating rate of 10 °C per minute in a nitrogen atmosphere with a flow rate of 200 mL per minute to study the thermal degradation of the composites.

Hardness test. The hardness of the pure ABS polymer as well as the prepared composites was measured by the Shore-D method using a Durometer hardness tester.

Results and discussion

Viscoelastic properties

DMTA characterization has been done to understand the viscoelastic behaviour of the prepared composites. Apart from the viscoelastic nature this test can also give information about the mechanical strength of the material, which could be related to the static mechanical properties.³

The energy stored per cycle via the material is given by the storage modulus (E′). The obtained results of E′ for the pure ABS as well as the ABS/GF composite with a varying GF content are shown in Fig. 1(a). It is observed that the addition of GFs in the ABS polymer matrix considerably enhances its E′, hence increasing the energy storage capacity of the composites. The composite filled up to 3 vol% of GFs shows less increase in E′ compared to the pure ABS polymer. Further addition of GF content increases E′ significantly. The nature of the curves are the same for all the compositions throughout the temperature range. There is a ∼1.6 times increment in the storage modulus of the composite consisting of 40 vol% GF to the pure ABS polymer. It is evident from the plot of E′ that the initial E′ for all the ABS/GF compositions is relatively high, then it starts decreasing with increasing temperature. As the temperature increased the weak interaction among the ABS chains diminished as shown in Fig. S1.† The possible interaction among the ABS and GFs could be weak interface interaction (van der Waals) due to the absence of chemical bonding.²⁶ Also, the increment in the temperature of the polymer matrix may reduce the stiffness of the composite during the application of heat since the bonding between the GFs and ABS weakened. Fig. 1(b) shows the variation of E′ calculated at four different temperatures for the composites having a varying GF content. E′ exhibits an increasing trend as the GF content increases in the composite. This is due to the fact that the addition of GFs allows the ABS/GF composite to absorb more and more energy in comparison with the pure ABS polymer up to the fixed deformation. The observed increment in the E′ value of the ABS/GF composite having 40 vol% GFs compared to pure ABS at room temperature is ∼250%, which remains almost the same for all other temperatures.

Fig. 1 (a) Storage modulus vs. temperature for 0 to 40 vol% loaded ABS/GF composites, and (b) change in storage modulus with GF content at four different temperatures.

The degree of entanglement between the polymer and filler in the polymeric composite has been also calculated from E′ obtained by DMTA. The following equation has been taken into consideration for the calculation of the degree of entanglement (N):^24,42

where R is the universal gas constant, T is the absolute temperature and E′ is the storage modulus.

Fig. 2(a) shows the variation in N for pure ABS as well as the ABS/GF composite having a varying percentage of GFs, calculated in the rubbery region at 110 °C. It has been observed that N increases as the filler content increases in the ABS/GF composites. At a lower loading of GF the degree of entanglement shows less improvement until 3 vol% addition of GFs. As the concentration of reinforcement increases the probability of interaction among the filler particles increases significantly; this may reduce the degree of entanglement but the behaviour of entanglement is increasing in nature, which suggests less interaction between filler particles. The trend is in agreement with the storage modulus (Fig. 1(b)).

Fig. 2 Variation in (a) degree of entanglement, (b) reinforcement efficiency, and (c) C factor for ABS/GF composites with varying GF content.

Apart from the degree of entanglement, information regarding the reinforcement efficiency factor can also be extracted with the help of DMTA. The reinforcement efficiency factor gives an idea regarding the filler matrix bonding by taking into consideration the effect of fractional addition of the filler into the polymeric composites. The calculation is performed using the Einstein equation:⁴³

E_c = E_m(1 + rV_f) (2)

where r is the reinforcement efficiency factor, V_f is the filler volume, E_c is the composites storage modulus and E_m is the matrix storage modulus. Fig. 2(b) shows the plot for the reinforcement efficiency factor of the ABS/GF composite systems having a varying GF concentration. It has been observed that the reinforcement efficiency factor shows low values for the composite having 1 and 3 vol% GFs and increases thereafter attaining maxima for a composite having 6 vol% GFs and further addition of GFs decreases the reinforcement efficiency. For a higher concentration of GFs the variation in reinforcement efficiency is not very significant, thus it can be concluded that the reinforcement efficiency is hugely dependent on the filler concentration and a better dispersion has been achieved for composites having 20 vol% GFs (with respect to the maximum loading).

The performance of the ABS/GF composites in the transition zone of E′ can be derived by the C factor. The C factor determines the effectiveness of the filler into the composite material and is given by the equation:⁴⁴

where E′_r and E′_g are the storage modulus values in the rubbery and glassy regions, respectively.⁴⁵ The C factor is also determined using the storage viscosity.⁴⁵ Fig. 2(c) shows the variation in the C factor determined for all compositions using E′. It can be easily seen that as the GF percentage increases, the difference in E′ for the rubbery and glassy regions goes down. The observed behaviour shows a higher value of the C factor for the composite having 1 vol% GFs and it keeps on decreasing as the GF content increases in the ABS/GF composites. The C value here gives information about the relative decrease in the E′ value with an increment in the temperature, which allows the material to pass through its T_g. The higher the value of C, the lower the effectiveness will be, and a lower proficiency of the distribution of fillers, which is the reason for the high C value, lowers the effectiveness of the ABS/GF composites.

The determination of the loss modulus (E′′) of the prepared composites is important to know the energy dissipation per unit cycle by the composite material. Fig. 3 shows E′′ for the pure ABS polymer to the composite consisting of 1 to 40 vol% GFs and the inset of Fig. 3 demonstrates the peaks of the E′′ curves. An increment in E′′ has been observed with an increasing GF content, which indicates an increase in the damping behaviour of the composite. Two significant peaks are visible in the plot of E′′ for the ABS/GF composite. These two peaks denote the presence of butadiene and styrene–acrylonitrile (SAN) phases of the ABS polymer.¹ The observation also shows that the peaks of these phases shift towards the higher temperature side as the content of GFs increases in the composite. The maximum shift observed for the SAN phase is 3.4 °C and for the butadiene phase it is 4.9 °C. The shift in the peak position also suggests the stability of the ABS/GF composite at higher temperature with an increase in the GF content. The increase in E′′ at low temperatures is due to the presence of the butadiene phase in the ABS polymer. The determined values of the ABS/GF composites at four different temperatures are shown in Fig. S2† with the varying GF percentage. The increment in the E′′ values from pure ABS to the composites with 3 vol% is less, which increases with the increasing filler content. From the plot (Fig. S2†) it is also clear that even around room temperature the effect on E′′ can be easily noticed. E′′ increases significantly as the GF content increases from pure ABS to the composite having 9 vol% GFs then from 12 vol% GFs to 40 vol% GFs in the ABS/GF composite. This behaviour is similar to that observed in the case of E′. It is evident from the results that the peak shifts towards the right, which concludes that the addition of GFs in the ABS matrix enhances its thermal stability. The possible reason for this could be the stability of GFs at very high temperatures as compared to the pure ABS polymer.

Fig. 3 Loss modulus vs. temperature for 0 to 40 vol% of GFs in the ABS/GF composites (inset shows the enlarged view of the loss modulus peak with temperature).

The damping factor (tan δ) ​can be defined as the ratio of E′′ to E′. Fig. S3† shows the variation of tan δ with temperature for the pure ABS polymer to the ABS/GF composite having 0 to 40 vol% GFs. To get an idea about the behaviour of peaks, the inset of Fig. S3† shows the enlarged view of the tan δ peaks. These plots provide information regarding the T_g and damping factor (peak value of tan δ at T_g). The obtained results of E′ and E′′ have already shown an increasing trend. It is also evident from the inset of Fig. S3† that the height of the tan δ peak decreases as the GF content increases in the composite. Apart from this, the graph also indicates a shift in peak positions toward a higher temperature. The ability of GFs to withstand high temperature could be the reason for this shift. The shift in the peak position (5.5 °C) towards the higher temperature region suggests that the ABS/GF composite’s thermal stability increases compared to that of the pure ABS polymer as the concentration of GFs increases.¹⁸

Fig. 4 shows the variation in the T_g and peak height of the composites containing varying % of GFs. The T_g shows an increasing trend as the content of GFs increases in the composites. The T_g for the pure ABS is 117.3 °C, which increases up to 122.8 °C for the composite having 40 vol% GF. It has been seen that the tan δ peak height for pure ABS is 2.68, which keeps on decreasing and finally settles at 1.88 for the composite containing 40 vol% of GF. The decreasing tan δ with more and more addition of GFs suggests that the storage property of the composite is dominant over the loss property.

Fig. 4 T_g and peak height vs. GF content in the ABS/GF composites.

The adhesion factor (A) is calculated with the help of damping factor values of the ABS polymer and ABS/GF composites. The volume fraction of the reinforcing filler has the key role in the determination of A. The simplified equation for the determination of A can be represented as:⁴⁶

where A is the adhesion factor, and tan δ_c and tan δ_p are the damping factors of the composite and polymer matrix, respectively. ϕ_β is the volume fraction of the filler. Fig. 5(a) shows the determined values of A for the ABS/GF composite with a varying GF content. The observation reveals that A keeps on decreasing with the increasing filler concentration until 20 vol% GF addition, and further inclusion of GFs increases A. The adhesion factor is dependent on the tan δ_c/tan δ_p ratio. The high volume fraction of filler and lower damping ratios of the composite to polymer for 20 vol% GFs decrease the adhesion factor. As the concentration of filler increases the possible interaction between GFs should increase which will lower the value of A. Jyoti et al. reported a similar observation for a MWCNT-reinforced ABS composite.²⁴

Fig. 5 (a) Adhesion factor for the ABS/GF composites with varying GF content and (b) variation in the b factor and tan δ_c/tan δ_p for the ABS/GF composites with varying GF content.

DMTA can be used to determine the property of the interphase between the polymer matrix and the reinforcing filler. The dissipation in the polymeric composites is not only due to the polymer matrix but also it has an effect on the interaction between the filler and matrix. The dissipation by the filler is dependent on the matrix, which tends to restrain the interface between the filler and matrix. The calculation can be performed using the relation:²⁴

where tan δ_c is the damping parameter for the composite and tan δ_p is for matrix. V_r is the fractional volume of filler and b is the exact volume occupied by reinforcement. Fig. 5(b) represents the variation of the b factor and tan δ_c/tan δ_p for the varying GF content. It can be seen that the behaviour of both factors is similar and exhibits a decreasing trend with the increasing filler concentration in the ABS/GF polymer composites. The damping ratios have not changed much after the composite having 20 vol% GFs thus it can be taken as the composite with a better dispersion of fillers. The b factor is the parameter that gives correctness in the reinforcement volume fraction due to the formation of a restraining interphase layer, and it should be decreasing in nature with the filler addition. The reason for this could be the increase in the interaction between the GF particles among themselves at a higher GF loading around T_g, and a similar behaviour has been shown by the ABS/GF composites.

The Cole–Cole method is a way to study changes introduced by the reinforced filler into the structural properties of the polymeric matrix. In this method the information regarding the dielectric relaxation is obtained via plotting each point of loss and storage moduli at the same frequency. Fig. 6 shows the Cole–Cole plot for the ABS/GF composite systems having a varying percentage of GFs, plotted at a frequency of 1 Hz. The Cole–Cole curve also tells about the nature of the composite system as obtained via damping behaviour. It has been reported that the shape of the curves (semicircle or elliptical) define the homogeneity of the system.²⁴ The analysis of the Cole–Cole curves shows that the nature of the curves is elliptical. The elliptical or imperfect circular nature of the curve denotes the very good adhesion between the filler and the polymer matrix.

Fig. 6 Cole–Cole plot for the pure ABS and ABS/GF composites.

The rheological properties of the polymer melt change with addition of different filler materials. These are affected by the size, shape and the concentration of the fillers and reinforcements.¹ The melt viscosity variation for the different composition with time is shown in Fig. S4.† The obtained melt viscosity of the ABS/GF composites exhibits an increasing behaviour as the GF concentration increases in the polymer matrix. This increment can be due to the hindrance offered by GFs in the flow of the polymer matrix. At the processing temperature of 230 °C the polymer matrix ABS is in the melt form while the filler GFs is in the solid form thus offering a large resistance to the flow of the molten mixture.

From the graph as shown in Fig. S4, † it is evident that the melt viscosity of the pure ABS is much lower than that of the ABS/GF composite containing 40 vol% of GFs. While processing pure ABS, the restriction in flow is considered only due to its long molecular chains, and thus the polymer melt shows lower viscosity. As the concentration of GFs increases, the shear stress of the molten mixture increases resulting in the increase in the melt viscosity. Fig. S4† also shows the variation of melt viscosity with processing time for a particular composition. There is a slight decrease in the melt viscosity. Once the feeding of the material is complete, the processing time starts, thus at the start of the processing it shows a slightly higher value since the restriction offered by the GFs in the flow of the mixture is more, but as the time passes the mixing of ABS and GF becomes more and more uniform thus offering less restriction in the flow of the melt mix resulting in slight decrease in the viscosity.

Complex viscosity results are also obtained from the DMTA analysis. Fig. 7 shows the complex viscosity plot for the pure ABS polymer to composite having 0 to 40 vol% of GFs. The complex viscosities show an increasing trend as the GF content increases in the ABS/GF composite. GFs have a very high melting point and the increase in temperature does not affect their structure. Meanwhile, an increase in temperature easily affects the ABS polymer, which easily changes its state from glassy to rubbery to liquid as the temperature increases. The behaviour of complex viscosity is the same as the melt viscosity and both show an increasing trend for the ABS/GF composite. The humps near −82 and 109 °C are due to the presence of butadiene and SAN phases in the ABS polymer.

Fig. 7 Variation of complex viscosity with temperature for different ABS/GF composites.

DMTA data are also used for the calculation of storage viscosity of the ABS/GF composites.^45,47 The elastic part of the complex viscosity is known as the storage viscosity. The expression for storage viscosity is:

where E′ is the storage modulus, ω is the frequency and ν is Poisson’s ratio. The loss viscosity has also been determined. Fig. 8(a) and (b) show the plot for the behaviour shown by the storage viscosity and loss viscosity, respectively with respect to temperature for the varying GF content in the ABS/GF composites. It has been noticed that with the increment in the GF vol% in the ABS matrix the storage as well as the loss viscosity increase. The storage viscosity starts from a higher value and decreases with the increment in the temperature as depicted by Fig. 8(a). The decrease is in two steps corresponding to the butadiene and SAN phases present in ABS, and overall the trend followed by the storage viscosity is similar to E′. The observed behaviour of loss viscosity is also increasing in nature with an increment in the GF concentration. From Fig. 8(b) the presence of two phases is confirmed by the two peaks appearing in the glassy and rubbery region. The observed nature of the plot for the loss viscosity is fairly similar to E′′.

Fig. 8 Behaviour of (a) storage viscosity, and (b) loss viscosity with temperature for the ABS/GF composites having varying GF content.

The viscosity trend observed for the ABS/GF composites using various factors (melt viscosity, complex viscosity, storage viscosity and loss viscosity) have shown an increasing trend as the concentration of GFs increases in the composites. With the increasing temperature the storage viscosity decreases, and the complex and loss viscosities also decrease but show peaks related to the two phases present in the polymer matrix. The melt viscosity measured while processing with time increases with filler concentration but decreases slightly as the time progresses for each composition. Overall all the viscosities show an enhancement with reinforcement of GFs.

The storage viscosity values are also used for the calculations of degree of entanglement, reinforcement factor and C factor by replacing E′ values with storage viscosity in eqn (1)–(3), respectively. Fig. 2(a) also shows the variation in N for pure ABS as well as the ABS/GF composite having a varying percentage of GFs, calculated in the rubbery region at 110 °C. It has been observed that N increases as the filler content increases in the ABS/GF composites. The observed behaviour of N calculated by E′ and storage viscosity is identical. The reinforcement efficiency factor is also calculated with the help of storage viscosity. Fig. 2(b) also shows the plot for the reinforcement efficiency factor of the ABS/GF composite systems having a varying GF concentration. The nature is similar to that of E′. The C factor is also determined using storage viscosity. Fig. 2(c) also shows the variation in the C factor determined for all compositions. The observations also reveal that the calculated values of the C factor from E′ as well as the storage viscosity follow the same trend.

Morphology

The morphology obtained via SEM of the plane ABS and fractured surface of the composites via the tensile test is shown in Fig. 9(a)–(d) and Fig. S5(a)–(f).† Fig. 9(a) shows the morphology of the plane ABS surface while Fig. 9(b) shows the fractured ABS surface. Subsequently Fig. 9(c) and (d) consist of the morphologies of the fractured surfaces of 20 and 40 vol% GF-containing ABS/GF composites. Micrographs of composites having 1 to 12 and 30 vol% GFs are shown in Fig. S5† in the increasing order of filler. The micrographs show the existence of smaller as well as bigger GF particles in the ABS matrix. The presence of GF particles in a smaller size is due to the breaking of GFs during compounding in the micro-compounder. The small pieces of GFs help in the better dispersion of the filler into the polymer matrix, which is responsible for the improvement in the homogeneity of the composites. However, the distribution of GFs is uneven at a higher loading of GFs. Furthermore, the increasing filler content increases the close association between the GF particles into the ABS matrix, which tends to increase the density of the composites, which is also responsible for the improvement in thermal conductivity. The decreasing distance among the GF particles allows them to contact one another, which leads to the formation of high thermal conductivity channels, enhancing the thermal conductivity of the composites.⁴⁸ On a closer look at the fractured surface of the composites, it can be also noticed that at very high filler loading the very small particles of the GFs accumulate at random places (during the final sample preparation via injection moulding) which may be responsible for the failure at a lower strain value of the composite during mechanical testing suggesting an increment in the brittleness of the ABS/GF composites.

Fig. 9 SEM micrograph of the ABS/GF composites: (a) plain ABS surface and (b) fractured ABS surface; fractured surfaces of the composites having (c) 20 and (d) 40 vol% GFs.

Mechanical properties

Tensile test is the way to understand the deformation behaviour and mechanical strength of the composite samples under uniaxial load. Also the processing parameters of moulding have a tendency to influence the properties of the final product.⁴⁹ In this study the processing parameters are kept constant in all composites. It has been observed that the ultimate tensile strength (UTS) of ABS/GF composites is a function of GF concentration, which is evident from the UTS plot for the ABS/GF composites in Fig. 10. The size of the GFs is quite high (∼150 micron) in comparison with the fillers used to prepare the composites of ∼20 μm.²⁸ Fig. 10 shows that the plastic deformation, which is shown in pure ABS starts to decrease as the composites have more and more GFs.²⁶ The pure ABS polymer shows the UTS of 51 MPa, which keeps on decreasing with further addition of GFs. The decrease in the UTS until 9 vol% is prominent as compared to the decrease in the UTS from 9 to 40 vol%. The decrease in the tensile strength can be attributed to the restricting effect of GFs on the polymer chain mobility under tensile loading, which in turn is also responsible for the brittle nature of the polymer composites.

Fig. 10 Variation in ultimate tensile strength (UTS), Young’s modulus and strain at break for varying vol% of GFs in the ABS/GF composites.

The strain at break of the ABS/GF composite exhibits a decreasing trend with the increasing content of GFs in the composite (Fig. 10). The observation shows a decrease of 38% in the fracture strain from pure ABS to the composite with 6 vol% GFs. From 6 to 40 vol% of GFs there is a nominal decrease in the strain at break. Fig. 10 also consists of Young’s modulus for the ABS/GF composites for varying vol% of GFs in the ABS/GF composites. The modulus exhibits an initial increase until 6 vol% of GF content; thereafter it stars decreasing. As compared to the pure ABS polymer, the composite containing 6 vol% GFs shows a 52 MPa increase in the modulus. A sharp decrease in the modulus value has been observed from 6 to 12 vol% GF reinforcement. Although the moduli are decreasing still the 30 and 40 vol% GF-reinforced composites are at the higher side with respect to pure ABS. Akinci et al. also reported a similar conclusion while studying polypropylene composites reinforced with graphite.⁵⁰

To know the performance of the ABS/GF composites under stress, a three-point bending test has been performed. Fig. 11(a) represents the maximum bending stress determined via a flexural test as well as the stress–strain behaviour under three-point bending for all the ABS/GF composites. The maximum bending stress decreases with increasing the GF content in the ABS/GF composites. It decreases sharply up to the 12 vol% GF content in the matrix and after that it decreases slowly. The flexural stress for the pure ABS polymer is observed to be 83 MPa. Addition of GFs decreases the flexural stress and finally it settles down to 67 MPa for the composite with 40 vol% GFs. This behaviour can be seen in the inset of Fig. 11(a), where the stress and strain value lowered for increasing the GF content in ABS. It may be as a result of an increase in the brittle nature and weak bonding between the reinforcement and matrix.

Fig. 11 (a) Maximum bending stress under three-point bending for the ABS/GF composites, and (inset) stress–strain behaviour of the ABS/GF composites under three-point bending, and (b) variation of the flexural modulus, flexural strain and impact strength with varying vol% of GFs.

Fig. 11(b) shows the variation of flexural modulus and flexural strain with respect to varying the GF content in the ABS/GF composites. The ABS polymer shows high flexural strain, which keeps on decreasing as the GF proportion increases in the composites. The prominent decrease in flexural strain is observed between pure ABS polymers to the composite containing 1 vol% GFs. The decrease in the flexural strain suggests that the material is becoming more and more brittle with the increasing GF content. The flexural modulus for the pure ABS polymer is observed to be 2.5 GPa. Addition of GFs increases the flexural modulus slowly until 9 vol% GF addition, and finally it reached 4.9 GPa for the composite with a 40 vol% GF content. Thus a trend of enhancement has been followed by the flexural modulus for the ABS/GF composites as the reinforcing filler GFs increases in the ABS polymer matrix. The composite with 40% GFs shows a 93% increment in the flexural modulus as compared to that of pure ABS. The increment in the flexural modulus could be attributed to the increasing hardness of the composites.

Fig. 11(b) also contains the plot for the variation in impact strength with the varying GF content. It is found that the impact strength of the ABS/GF composites decreases with the addition of GFs in ABS. A sudden drop has been observed in the impact strength from 88 to 52 J m⁻¹ (41%), on the addition of only 1 vol% GFs in ABS. The impact strength has decreased with a lower rate on further addition of GFs up to 6 vol% in the ABS matrix. Thereafter it has decreased with a small magnitude on further addition of GFs (up to 40 vol%). It may be assumed that due to the addition of highly stiff GFs in the ABS matrix, the mechanism of energy absorption has changed. However, the reduction in impact strength is normally observed on the incorporation of rigid fillers into a comparatively tough polymer matrix. Similar results have been observed by Pour et al. in a study of a graphene-reinforced ABS/PC nanocomposite.⁴⁰

Thermal conductivity

Thermal conductivity is a materials property, which helps in the transport of heat energy. Heat transfer through a polymer composite depends on the interfacial thermal conductance of the polymer matrix and filler.¹ The main issue in the case of a polymer is low interfacial thermal conductance.

Fig. 12 shows the variation of thermal conductivity of the pure ABS polymer and composites containing 12 to 40 vol% of GFs. Due to the presence of a lower amount of GFs, no significant changes are observed in the thermal conductivity of the pure ABS polymer to the composite containing 1 to 9 vol% of GFs. Thus to make clear distinctions, the thermal conductivity for the composites containing 1 to 9 vol% GFs is not shown here. The thermal conductivity of the ABS/GF composite shows an increasing behaviour in comparison to that of the pure ABS polymer. The higher thermal conductivity of the GFs as compared to the ABS polymer is the reason for this increase. As the GF concentration increases, the number of layers or phases of GFs increases, which increases the transport of thermal energy through the ABS/GF interface. The interface is transporting more and more thermal energy, causing an increase in the thermal conductivity of the ABS/GF composites.⁴⁸

Fig. 12 Thermal conductivity for 0 to 40 vol% of GF-reinforced ABS/GF composites.

Thermogravimetric analysis (TGA)

TGA of the ABS/GF composites is shown in Fig. 13. Thermograms of weight loss vs. temperature indicate clearly that the ABS/GF composites have a slightly better thermal stability than the pure ABS. It has also been observed that the decomposition temperature of the composite has increased with increasing the vol% of the GF loading in the ABS matrix. The higher shift in degradation temperature has been observed at a 40 vol% GF loading. To investigate the effect of the GF loading on the degradation behaviour of ABS, the inset of Fig. 13 contains an enlarged view from 330 to 392 °C. The apparent drying of the material, gas purging rate, sample weight and also the heating rate affect the degradation of the composite material.⁵¹ The change in the onset temperature is ~2.9 °C. Al-Saleh et al. reported a similar observation for CNT addition in ABS.⁵² This may also happen as the interaction is physical (van der Waals interaction). It can be seen that the char yields after degradation of the composites have increased from 2% (pure ABS) at 480 °C to 35% (40 vol% GFs in ABS) on increasing the addition of the GF content in the composites at the same temperature. This enhancement in char yields and thermal stability of the graphite-reinforced composites can be attributed to interfacial interactions between the poor thermally stable ABS with highly thermally stable GFs. This can also result from the formation of a layer network during the combustion of the composite, which leads to inhibition of the diffusion of gaseous products out after decomposition of the composites. Similar results have been observed by Pour et al. in the case of a graphene-reinforced PC/ABS nanocomposite.⁴⁰

Fig. 13 Weight loss (%) vs. temperature for 0, 3, 6 and 40 vol% GFs in the ABS/GF composites.

Hardness and density

Fig. S6† shows the hardness of the composites with an increasing amount of GFs. The hardness for the pure ABS polymer is 81, which increases up to 83 for the 6 vol% GF-filled composite. This shows that the ABS matrix has a dominant role for the composites until 6 vol% of GFs. After this the hardness value starts to decrease slightly until the composite with 40 vol% GFs. The reason for the decrease in the hardness can be attributed to the presence of fewer adhesive forces between the constituents of the composite.²⁸

The variation in density of the ABS/GF composites is also illustrated in Fig. S6† with the varying vol% of GFs. The density of the ABS polymer is 1.06 g cm⁻³. It is evident from the graph that on inclusion of filler particles up to 6 vol%, there is a nominal increase in the bulk density of the composite. Due to the lower amount of GFs (until 6 vol%) in the composite, the ABS matrix dominates the bulk density and restricts its increment. As we keep on further increasing the GF content, the increment in density becomes more and more prominent, since the bulk density of graphite is higher than the density of the ABS polymer.² For the composite with 40 vol% of GFs, the observed increase in the bulk density is 24% in comparison with the pure ABS polymer.

To compare the mechanical behaviour of the prepared ABS/GF composites with the previously reported ABS composites having different fillers, normalisation of the modulus has been done. The following equation has been applied for this purpose:

where χ_i represents the value for the polymer matrix, χ_γ is the value of the composite with filler and W_γ is the weight percentage of the incorporated filler.

Apart from the modulus, eqn (7) has also been used for the comparison of T_g obtained via DMTA. Fig. 14(a) and (b) show the comparison of the modulus and T_g for ABS composites having different fillers. The obtained results show that the modulus and T_g of the ABS/GF composite system are comparable with those of the ABS composites reinforced with different fillers.

Fig. 14 (a) Percentage enhancement in the modulus on addition of different fillers (graphite,²⁶ hollow glass beads,⁹ TiO₂,⁴ alumina,¹⁰ CB blend,²⁰ wood dust,³⁶ CaCO₃,⁷ ZnO⁵ and CNT blend²⁰) in the ABS polymer. (b) Percentage enhancement in T_g obtained via tan δ (DMTA) on different filler (resin,²⁵ CNT/PS,¹⁸ PS,²¹ cotton,³⁴ iron powder³⁵ and CNTs¹⁸) addition in the ABS polymer.

Conclusions

Composites consisting of ABS as the matrix and different vol% of GFs as the filler material were successfully prepared using melt blending by a micro-compounder having vertical twin screws in co-rotating fashion and the injection moulding technique. During compounding it is observed that the melt viscosity of the composites keeps on increasing as the GF content increases, and the complex viscosity confirms the same. The density of the composite increases as the GF vol% increases, which is also the reason for the increase in thermal conductivity. The storage as well as the loss moduli show the increasing behaviour as the GF concentration increases. A shift in the tan δ peak towards a higher temperature range suggests an increment in the thermal stability of the composites. This result is confirmed by TGA analysis, which shows a shift in the degradation temperature on addition of GFs at a higher level in the ABS matrix. DMTA has also been used to study the degree of entanglement, reinforcement efficiency and C factor, showing better adhesion between the polymer and filler. The hardness of the composite shows a mixed behaviour, which increases for the lower concentration of GFs and then decreases for a higher GF concentration. The UTS of the composites shows a significant decrease, whereas the brittleness increases as we keep on adding GFs into the ABS polymer, which is also confirmed by SEM. The impact test shows a significant reduction in impact strength even on addition of 1 vol% GFs in the polymer. The bending stress and strain decrease while the flexural modulus shows an increasing trend with the increasing GF content in the composite. The prepared composite shows a significant improvement in the thermal and some of the mechanical properties and may be used for making electronics and electrical devices including bipolar plates for fuel cell applications.

Acknowledgements

Authors acknowledge the financial support provided by Indian Space Research Organization, Government of India for carrying out this research work.

References

1. J. A. Brydson, Plastics Materials, Butterworth Heinemann, Oxford, 7th edn, 1999.
2. K. K. Kar and A. Hodzic, Development in nanocomposite, Research Publishing Services, Singapore, 2014.
3. K. K. Kar, J. K. Pandey and S. K. Rana, Handbook of Polymer Nanocomposites. Processing, Performance and Application: Volume B: Carbon Nanotube Based Polymer Composites, Springer, Heidelberg, New York, Dordrecht, 2014.
4. J. Sudeepan, K. Kumar, T. Barman and P. Sahoo, J. Inst. Eng. (India): Ser. C, 2015, 1–13,  DOI:10.1007/s40032-015-0192-y.
5. J. Sudeepan, K. Kumar, T. Barman and P. Sahoo, Mater. Manuf. Processes, 2015, 5, 1–11.
6. C. Y. Tang, L. C. Chan, J. Z. Liang, K. W. E. Cheng and T. L. Wong, J. Reinf. Plast. Compos., 2002, 21, 1337–1345.
7. L. Jiang, Y. C. Lam, K. C. Tam, T. H. Chua, G. W. Sim and L. S. Ang, Polymer, 2005, 46, 243–252.
8. A. Zhang, G. Zhao, Y. Guan and G. Chen, Polym. Compos., 2010, 31, 1593–1602.
9. J.-Z. Liang, J. Elastomers Plast., 2005, 37, 361–370.
10. K. K. Kar, S. Srivastava, A. Rahaman and S. K. Nayak, Polym. Compos., 2008, 29, 489–499.
11. S. Rai and S. P. Priya, J. Reinf. Plast. Compos., 2006, 25, 565–574.
12. G. Yan, X. Wang and D. Wu, J. Appl. Polym. Sci., 2013, 129, 3502–3511.
13. A. Shenavar and F. Abbasi, J. Appl. Polym. Sci., 2007, 105, 2236–2244.
14. R. K. Mondal, K. A. Dubey, Y. K. Bhardwaj, L. Panicker and L. Varshney, Polym. Compos., 2016, 37, 481–487.
15. D. Mari and R. Schaller, Mater. Sci. Eng., A, 2009, 521–522, 255–258.
16. N. K. Shrivastava, S. Suin, S. Maiti and B. B. Khatua, RSC Adv., 2014, 4, 24584–24593.
17. Y. Sun, Z.-X. Guo and J. Yu, Macromol. Mater. Eng., 2010, 295, 263–268.
18. B. K. Singh, P. Kar, N. K. Shrivastava, S. Banerjee and B. B. Khatua, J. Appl. Polym. Sci., 2012, 124, 3165–3174.
19. M. C. L. De Oliveira, G. Ett and R. A. Antunes, J. Power Sources, 2013, 221, 345–355.
20. O. Meincke, D. Kaempfer, H. Weickmann, C. Friedrich, M. Vathauer and H. Warth, Polymer, 2004, 45, 739–748.
21. E. M. Araujo, E. Hage Jr and A. J. F. Carvalho, J. Mater. Sci., 2004, 39, 1173–1178.
22. A. Rostami, M. Masoomi, M. J. Fayazi and M. Vahdati, RSC Adv., 2015, 5, 32880–32890.
23. W.-Y. Wang, G.-H. Luo, F. Wei and J. Luo, Polym. Eng. Sci., 2009, 49, 2144–2149.
24. J. Jyoti, B. P. Singh, A. K. Arya and S. R. Dhakate, RSC Adv., 2016, 6, 3997–4006.
25. P. Jyotishkumar, E. Abraham, S. M. George, E. Elias, J. Pionteck, P. Moldenaers and S. Thomas, J. Appl. Polym. Sci., 2013, 127, 3093–3103.
26. B. Ben Difallah, M. Kharrat, M. Dammak and G. Monteil, Mater. Des., 2012, 34, 782–787.
27. F. M. Uhl, Q. Yao and C. A. Wilkie, Polym. Adv. Technol., 2005, 16, 533–540.
28. H. S. Dahiya, N. Kishore and R. M. Mehra, J. Appl. Polym. Sci., 2007, 106, 2101–2110.
29. L. L. Wang, L. Q. Zhang and M. Tian, Mater. Des., 2012, 39, 450–457.
30. S. K. Sharma, V. Gupta, R. P. Tandon and V. K. Sachdev, RSC Adv., 2016, 6, 18257–18265.
31. M. Mariatti, M. Nasir and H. Ismail, J. Reinf. Plast. Compos., 2002, 21, 711–722.
32. S.-C. Chen, M.-C. Lin, R. Der Chien, M.-Y. Teng and C.-S. Chen, J. Reinf. Plast. Compos., 2004, 23, 1115–1129.
33. S. R. Owen and J. F. Harper, Polym. Degrad. Stab., 1999, 64, 449–455.
34. J. N. Martins, T. G. Klohn, O. Bianchi, R. Fiorio and E. Freire, Polym. Bull., 2010, 64, 497–510.
35. M. Nikzad, S. H. Masood and I. Sbarski, Mater. Des., 2011, 3448–3456,  DOI:10.1016/j.matdes.2011.01.056.
36. L. Chotirat, K. Chaochanchaikul and N. Sombatsompop, Int. J. Adhes. Adhes., 2007, 27, 669–678.
37. X. Ma, H. Ge and H. Liu, Polym. Compos., 2015 DOI:10.1002/pc.23491.
38. R. Mujal-Rosas, J. Orrit-Prat, X. Ramis-Juan, M. Marin-Genesca and A. Rahhali, Adv. Polym. Technol., 2013, 32, E399–E415.
39. S. Bonda, S. Mohanty and S. K. Nayak, Polym. Compos., 2012, 33, 22–34.
40. R. H. Pour, A. Hassan, M. Soheilmoghaddam and H. C. Bidsorkhi, Polym. Compos., 2016, 37, 1633–1640.
41. R. K. Gautam, S. Banerjee and K. K. Kar, Recent Pat. Mater. Sci., 2015, 8, 15–45.
42. Z. Oommen, G. Groeninckx and S. Thomas, J. Polym. Sci., Part B: Polym. Phys., 2000, 38, 525–536.
43. A. Einstein, Investigations on the Theory of the Brownian Movement, Courier Corporation, 1956.
44. L. A. Pothan, Z. Oommen and S. Thomas, Compos. Sci. Technol., 2003, 63, 283–293.
45. C. W. Macosko, Rhelogy-principle, measurements, and applications, Wiley-VCH, New York, 1994.
46. C. Correa, C. Razzino and E. Hage, J. Thermoplast. Compos. Mater., 2007, 20, 323–339.
47. K. P. Menard, Dynamic mechanical analysis: a practical introduction, CRC press, 2008.
48. W. Yu, H. Xie, L. Chen, M. Wang and W. Wang, Polym. Compos., 2015 DOI:10.1002/pc.23802.
49. C. Chen, J. Reinf. Plast. Compos., 2006, 25, 1367–1373.
50. A. Akinci, Archives of Materials Science and Engineering, 2009, 35, 91–94.
51. S. Yang, J. Rafael Castilleja, E. V. Barrera and K. Lozano, Polym. Degrad. Stab., 2004, 83, 383–388.
52. M. H. Al-Saleh, B. A. Al-Saidi and R. M. Al-Zoubi, Polymer, 2016, 89, 12–17.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra09236e

---