Mechanical reinforcement with enhanced electrical and heat conduction of epoxy resin by polyaniline and graphene nanoplatelets

2.1 Materials

Epoxy resin comprises a group of bisphenol A and epichlorohydrin with an equivalent weight of 184–190 and a density of 1.16 kg/L, and polyetheramine D-230 as a hardener was supplied from Asachem (Malaysia) Sdn. Bhd. Needle-like PANI black green powder (purity: >99 wt%, yield: 40 wt%, conductivity: 4 W/cm²; average diameter: 8 μm; length: 100 Å; and electrical conductivity: 2.7 S/cm) was supplied by E-TEK Co. Ltd., Korea. The GNPs (purity: >99.5 wt%; diameter: 5–10 nm; thickness: 4–20 nm; density: 0.23 g/cm³, layers: <30; volume resistivity: 4 × 10⁻⁴ ohm cm) were supplied by Chengdu Organic Chemicals Co., Ltd. Chinese Academy of Sciences.

2.2 Sample preparation

Six samples with different formulation were prepared as described in Table 1. The schematic diagram of the preparation of epoxy, epoxy/PANI, and epoxy/PANI/GNPs nanocomposites is presented in Figure 1. For the three-component system, the 3 wt% PANI powders were first mixed with different loadings of GNPs; then, the PANI/GNPs mixture was mixed with epoxy with a mechanical stirring machine for 1 h at 750 rpm. To avoid the formation of aggregates, the mixtures were sonicated using a probe sonicator (Model: Hielscher UP400S, 400 watts, 24 kHz) for 3 h. During sonication, the mixture container was held in an ice bath to avoid overheating the suspension so that the temperature was maintained at 40°C. After sonication, the hardener (polyethylene D230) was added into the mixture at a ratio of 3:1, and the mixture was gently stirred for 5 min. Then, the mixture was pumped under vacuum for 10 min to eliminate any trapped air. In the final stage, all samples were cured for 24 h at room temperature and subsequently post-cured for 2 h at 80°C and 1 h at 110°C to stabilize and remove air bubbles.

Figure 1

Schematic diagram of the epoxy/PANI/GNPs melt-blending process.

2.3 Characterization

A bending test was performed with a three-point flexural test machine (Testometric M350-10CT model) based on the ASTM D790 standard, a load cell of 1,000 N and a cross-head speed of 1.37 mm/min. The dimensions of the specimens in this study were 127.0 × 12.7 × 3.0 mm³. The values were taken from an average of six samples.

The Izod impact strength test was performed at a load weight of 0.898 kg and a velocity of 3.45 m/s using a Ray–Ran Pendulum Impact System in accordance with ASTM D 256-90b. The sample dimensions were 63.0 × 12.7 × 3.0 mm³. The average value from six replicates was reported for each composition.

Differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) were run using a Mettler Toledo DSC 882^e and a TGA/SDTA851^e, respectively, with 15–20 mg of samples. The DSC samples were heated from 30 to 200°C at a rate of 10°C/min and subsequently cooled at the same rate under atmospheric air flow condition. The samples for TGA were tested from 30 to 600°C at a heating rate of 10°C/min.

The electrical conductivity of the samples was measured using a high-frequency response analyser (HFRA Solartron 1256; Schlumberger) from 0.1 Hz to 1 MHz. The sample was inserted between two stainless steel electrodes with a surface contact area of 2 cm² and subsequently mounted onto the holder. The bulk resistance (R _b (Ω)) of the samples was determined from the obtained Cole–Cole plots using Z-View software. The conductivity was determined according to equation (1):

where L denotes the thickness of the sample (cm), and A denotes the effective contact area (cm²) of the electrode and electrolyte.

The thermal conductivities of the samples (one specimen was tested per composite type) were measured using a thermal conductivity analyser (TCA, Nanoflash NETZSCH, model LFA 44712-41). Disc-type samples (with a diameter of 12.7 mm and a thickness of 1 mm) were put in an electric furnace. The thermal conductivity (λ, W m⁻¹ K⁻¹) of the samples, which depends on the thermal diffusivity (α, mm² s⁻¹), density (ρ, g cm⁻³), and specific heat capacity (C _p, J g⁻¹ K⁻¹), was calculated at various temperatures from 30 to 180°C by the following equations:

where d is the sample thickness (mm), t _1/2 is the time when T/T _max reaches 0.5, m is the sample mass, Q is the energy absorbed by the sample, and T _max is the maximum rear-side temperature increase. To calculate the heat capacity, pyroceram with the thickness of 12.7 mm was used as reference specimen.

The morphology of the samples was examined using high-resolution field emission scanning electron microscopy (FESEM, model SUPRA 55VP). The fracture surfaces of the impact test specimens were sputter coated with platinum before being studied by FESEM to avoid the charging effect and obtain a clear image.

3.1 Mechanical properties

The bending, tensile, and impact properties of neat epoxy, epoxy/PANI, and GNPs nanocomposites with different filler loadings are presented in Figure 2. In comparison to neat epoxy, the failure of the mechanical properties induced by the PANI component is primarily because of two causes. First, PANI acts as a non-reinforcing filler because of important differences in the crystallinity and polarity from the epoxy polymer; second, micro-voids and clusters form because of the enhanced viscosity of the composite [27].

Figure 2

Mechanical properties of epoxy, epoxy/PANI, and epoxy/PANI/GNPs nanocomposites.

The incorporation of GNPs significantly affected the flexural and tensile properties of nanocomposites, as indicated in Figure 2. The bending strength and bending modulus of pure epoxy improved with a 0.3 wt% loading of GNPs (NC2 nanocomposites) but subsequently gradually decreased with higher contents of GNPs. Compared to pure epoxy, the presented improvements in both bending strength and modulus of the NC2 nanocomposites were 14 and 25%, respectively. These improvements can be ascribed to the strong reinforcement effects of graphene, appropriate compatibility, and excellent interactions of the filler/epoxy matrix interface, which results in an efficient load distribution from the matrix to the filler [1]. However, with an additional increase in GNPs loadings, the bending properties of NC3 (0.5 wt%) and NC4 nanocomposites (0.7 wt%) deteriorated. This result is in agreement with previous studies on GO/epoxy by Chhetri et al. [28] and GNPs/epoxy composites by Shen et al. [29], which showed the optimal enhancement in flexural properties at 0.1 and 0.25 wt%, respectively.

As shown in Figure 2, the tensile properties revealed a trend similar to the flexural properties. Compared to pure epoxy, the tensile strength and modulus of the NC2 nanocomposite increased by 19 and 5%, respectively. These remarkable increases may be because GNPs as physical crosslinking areas limit the movement of the epoxy chains under a load. Similarly, the increase in tensile strength and Young’s modulus of epoxy by 23 and 10% were reported at 0.25 and 0.5 wt% GNPs, respectively, as reported by Shokrieh et al. [30]. Beyond the addition of 0.3 wt% GNPs, the tensile properties deteriorated with increasing GNPs loading. The reason is the nonuniform dispersion and agglomeration of GNPs/PANI at higher contents inside the epoxy matrix, where the packing behaviour of graphene arises from the high aspect ratio and van der Waals forces of GNPs. These agglomerates act as stress concentrators with the weakest areas at the centre points, which reduces the properties of the nanocomposites as described by Pour et al. who reported the lower mechanical strengths when the GO levels were more than 0.5 wt% in epoxy [31].

The strain of epoxy/PANI and NC1, NC2, NC3, and NC4 nanocomposites showed a continuous and significant reduction in Figure 2. This result can be ascribed to the increased rigidity of the epoxy matrix. The inserted GNPs/PANI fillers in the epoxy acted as limitation points for the polymer chain movement, which made the nanocomposites more brittle [32]. Nevertheless, the impact strength in Figure 2 displays the maximum enhancement in the NC2 nanocomposite because the homogeneous dispersion was restricted up to a GNPs concentration of 0.3 wt%. However, at higher GNPs loadings in NC3 and NC4, the significantly decreased impact strength might be because of the fillers agglomeration, which can produce unexpected internal deficiencies such as cracks to result in weaknesses and smaller energy absorption in the epoxy nanocomposite. The toughness enhancement is believed to depend on the stress concentration factor that is related to the size of graphene; for instance, 0.1 wt% GO was optimally obtained in epoxy [33], and 1.5 wt% [34] and 3 wt% GNPs [35] were achieved in TPNR.

3.2 DSC analysis

In this work, DSC was performed to test the thermal properties including measuring the glass transition temperature (T _g). T _g denotes the temperature range in which the thermoset polymer changes from hard, rigid, or “glass”-like to more flexible, submersible, or “elastic”. Based on Figures 3 and 4, the T _g of neat epoxy is located at 53.7°C, and the T _g of epoxy/PANI is at 53.4°C. Moreover, the analysis showed that the T _g shifted to higher values with the inclusion of GNPs/PANI in the epoxy matrix. The T _g of the nanocomposites increased by 1.1 and 3.3°C when the GNP content was 0.1 and 0.3 wt%, respectively. Increasing the weight percentage of GNPs from 0.5 to 0.7 wt% showed that the T _g decreased to 54.76 and 54.58°C, respectively, compared to nanocomposites containing 0.3 wt% GNPs. This trend occurs because the GNPs and PANI acted as physical interlocking points in the epoxy matrix, which limited the movement of epoxy chains. Moreover, the reduction in T _g might be caused by the decrease in crosslinking density or the extent of curing in the epoxy matrix. The increasing content of GNPs can induce the formation of aggregates that create physical defects, which obstructs the crosslinking of the epoxy matrix with the hardener and decreases the crosslinking density. For instance, Lin et al. found that the addition of PSS-PANI/rGO up to 1 wt% decreased the crosslinking density of epoxy matrix because of the filler agglomeration [25]. The development of aggregates also provides free volume at the interface, which increases the mobility of the segments. The decline in T _g of NC3 and NC4, as shown in Figures 3 and 4, could be caused by the restacking of GNPs sheets and PANI, which also form clusters in the matrix.

Figure 3

DSC thermograms of epoxy, epoxy/PANI, and epoxy/PANI/GNPs nanocomposites.

Figure 4

Glass transition temperature of epoxy, epoxy/PANI, and epoxy/PANI/GNPs nanocomposites.

3.3 Thermogravimetric analysis

Figure 5 displays the TGA and DTG curves to study the thermal stability of epoxy, epoxy/PANI, and epoxy/PANI/GNPs nanocomposite samples. The TGA data for all studied samples are described in Table 2. All investigated samples show a comparable thermal behaviour to DSC. Figure 5 shows that all samples exhibited a one-step decomposition process. For pure epoxy, the early decomposition began at a T _10% of 336.9°C and initially ended (which was recorded at a weight loss of 90%) at 445.4°C. By adding PANI into epoxy, the decomposition process was accelerated, as presented in the inset image in Figure 5, and the curves shifted to the left side. Moreover, the decomposition temperatures were lower (T _10% = 316.2°C) for the epoxy/PANI composite than those of the pure epoxy (Table 2). This result can be correlated with the initial decomposition of PANI at 153–295°C [36], where the same range of temperature for the pure PANI has been reported by Saritha and Sunil [37]. When GNPs were incorporated into epoxy/PANI, the rate of decomposition changed with the GNP loading. At a lower GNP loading (0.1 wt%) in NC1, the curve shifted to the left, and T _10% decreased to 306.2°C; however, the 90% decomposition point occurred at 462.5°C possibly may be because the nanofiller loading was too low and insufficient to cause a good distribution of GNPs. Interestingly, the TGA curves shifted back to higher temperatures with higher loadings of GNPs. At 0.3–0.7 wt% GNPs, the decomposition process was delayed (compared to that of epoxy/PANI), where T _10% were recorded at 321.1, 318.6, and 321.8°C, respectively, although they were still lower than that of pure epoxy. Thus, a higher thermal resistance occurred, which is a consequence of the effective interaction between GNPs and the epoxy because of the good dispersion and distribution of GNPs. In addition, the improved thermal stability can be correlated with the creation of a carbonaceous char structure during thermal oxidation [38].

Figure 5

TGA thermograms of epoxy, epoxy/PANI, and epoxy/PANI/GNPs nanocomposites.

The residue contents left after the decomposition process at 450°C are displayed in Table 2. The least amount of residual mass was found in pure epoxy (5.42%). Remarkably, when the GNPs and PANI were incorporated, the increase in residue amounts was 6.43–7.76%. The increased content of residue is ascribed to the existence of the PANI/GNPs, which is retained even after the decomposition of epoxy, which is consistent with the results of Rahnamol and Gopalakrishnan (2020) who incorporated 3 wt% PANI/GO hybrid in epoxy resin [22]. Moreover, GNPs can alter the thermal degradation path at high temperatures and endorse the formation of char from epoxy, as described in the previous study on epoxy/graphene nanosheets [39].

3.4 Electrical conductivity

The electrical conductivity of epoxy/PANI nanocomposites with various loadings of GNPs is presented in Table 3. Neat epoxy exhibited an electrical conductivity of 7.83 × 10⁻⁸ S/m, which implies an insulating behaviour; this result is consistent with the literature [40]. For the epoxy/PANI composite, the conductivity improved to 6.05 × 10⁻⁷ S/m at a 3 wt% loading. By adding GNPs, the conductivity of the NC1 (0.1 wt% GNPs) nanocomposite increased to 7.39 × 10⁻⁵ S/m, and the NC2 (0.3 wt% GNPs) nanocomposites showed a slightly higher conductivity of 8.72 × 10⁻⁵ S/m. This could be related to the better GNPs dispersion and interaction linkage of GNPs with epoxy/PANI as observed in Figure 8(d). In this condition, an effective conductive path can be created via GNPs layers with the PANI components to link the gap between GNPs layers and form a suitable percolating network structure in the matrix, which enables the electrons to pass through direct “contacts” between the conductive fillers [41,42,43], as illustrated in Figure 6. However, at higher loadings of GNPs, the electrical conductivity has poor performance. The epoxy nanocomposites with 0.5 wt% GNPs (NC3) and 0.7 wt% GNPs (NC4) had electrical conductivities of 2.44 × 10⁻⁵ and 8.09 × 10⁻⁶ S/m, respectively. This is possibly because the decreased interaction of the GNPs and PANI, which is caused by their random agglomeration along with a small aspect ratio of unity, makes it difficult in forming conductive networks in the insulating epoxy matrix [44].

Figure 6

Schematic diagram of the conductive path inside the epoxy/PANI/GNPs (NC2) nanocomposites.

3.5 Thermal conductivity

The thermal conductivity of epoxy, epoxy/PANI, and NC samples is displayed in Figure 7. The thermal conductivity of all samples increased with increasing temperature from 30 to 120°C. Furthermore, the effect of temperature on the thermal conductivity was insignificant when the testing temperature exceeded 120°C. NC2 had higher thermal conductivity than neat epoxy and the other samples at all measured temperatures. For instance, the increased thermal conductivity of NC2 at 30, 60, 90, and 120°C records for approximately 30, 31.3, 35.3, and 39%, respectively, compared to the pure epoxy. This improvement in thermal conductivity might be because the intrinsic thermal conductivity of the GNPs and PANI is much higher than that of pure epoxy [44]. Another possible reason for the increase in thermal conductivity is the homogeneous dispersion of GNPs/PANI fillers in the epoxy matrix which led to the increased interfacial area and the reduced interfacial thermal resistance of epoxy and GNPs/PANI. Yu et al. [45,46] noted that the low interfacial thermal resistance contributed by GNPs is ascribed to the decreased geometric contribution of phonon scattering at the interfaces. Furthermore, the GNPs/PANI components act as a phonon transport bridge and enable for the complete heat transfer within the epoxy nanocomposite. As observed in Figure 7, the thermal conductivity began to decrease with NC3 and NC4 because of the presence of agglomerated clusters and voids as strong thermal barriers, which significantly decrease the thermal conductivity of the nanocomposites [47,48,49].

Figure 7

Thermal conductivity of epoxy, epoxy/PANI, and epoxy/PANI/GNPs nanocomposites.

3.6 Morphological analysis

To examine the dispersion of hybrid PANI/GNPs in the epoxy matrix, the impact fractured surfaces of pure epoxy, epoxy/PANI, and GNPs filled nanocomposites were analysed. As displayed in Figure 8a, the surface of pure epoxy is smooth with various river patterns because of the extremely brittle type of epoxy. However, Figure 8b shows a relatively rough fractured surface of the epoxy/3 wt% PANI samples with holes, which appear to have a beehive shape. As shown in Figure 8c and d, the GNPs/PANI fillers were uniformly dispersed in the epoxy matrix without showing much agglomeration in the NC1 and NC2 samples. Especially in NC2 nanocomposite, Figure 8d displays a less extreme variation in the microstructure where the smooth/rough phase and distribution of holes are approximately identical in the micrograph. At low magnification, the image of the NC2 nanocomposites in Figure 8d shows the embedment of platelets in the holes or gaps of matrix and this confirms the presence of appropriate interactions and the good compatibility among the GNPs, PANI, and epoxy matrix. In addition, no obvious clusters of filler can be found, and the interactions seem to provide a network-like structure. This observation supports the significant improvement in the properties of the nanocomposites than that of pure epoxy. In the case of NC3 and NC4, as indicated in Figure 8e and f, many agglomerations of GNPs/PANI on the epoxy matrix surface are obvious, and the layered structure of GNPs is clearly visible in the NC3 and NC4 nanocomposites, as indicated by the black arrows. The agglomerates normally delaminate through the fracture process because of the weak adhesion between the fillers and the matrix. These agglomerates have an adverse effect on improving the performance of nanocomposites.

Figure 8

FESEM micrographs of (a) epoxy, (b) epoxy/PANI, (c) NC1, (d) NC2, (e) NC3, and (f) NC4.

According to the FESEM micrographs in Figure 8, a schematic model of the epoxy, epoxy/PANI, and epoxy/PANI/GNPs nanocomposites is presented in Figure 9. As a control sample, the epoxy system that acts as the polymeric matrix is represented by a smooth and one-phase surface. With the inclusion of 3 wt% PANI into the epoxy matrix, the added PANI with a needle-like dimension [50] (as represented by the rectangular shape) is expected to distribute well in the composite. The small amount of GNPs corresponds to an insufficient GNPs–PANI interaction in NC1 where the smooth and rough surfaces were obviously visible in different spot as shown in Figure 8c. In comparison, the distribution and dispersion of the GNPs (flaky GNP is presented by a crooked line) and PANI in the NC2 nanocomposite (which is consistent with Figure 8d) system show good polymer–filler contact because of the π–π interactions and hydrophobic–hydrophobic interactions [49,51], which effectively reinforces impact in the system, as previously highlighted. When the GNPs content is higher, such as in the NC3 and NC4 systems, the GNPs and PANI components experience intracellular localization, which promotes the formation of agglomerates in more clusters. For example, an inhomogeneous dispersion of GNPs is produced because of the increasing amount of GNPs in the system and increasing size of the PANI cluster.

Figure 9

Schematic diagram of epoxy, epoxy/PANI, and epoxy/GNPs/PANI nanocomposites.