3.1 Cure behavior and chemical interactions of ENR with nanofillers
Figure 1 illustrates the curing behaviors of gum ENR and its related compounds, which include ENR-GPx (where x represents different GP loadings) (Fig. 1(a)), ENR-CNT3, and ENR-CNT3/GPx hybrid composites with varying graphite loadings (Fig. 1(b)). All ENR compounds display reversion curves, characterized by an initial increase and subsequent decrease in torque after reaching the maximum value within a curing time period of approximately 5 to 15 min. This behavior is primarily attributed to the breakdown of weak chemical linkages within the ENR molecular networks. It is important to note that ENR is a reactive polymer containing unsaturated double bonds and epoxy sites, which are available for cross-linking reactions [33].
The high reactivity of the epoxirane rings and double bonds in ENR molecules enables both self-crosslinking among polar epoxirane rings and sulfur curing of ENR molecules, respectively. Sulfur-cured ENR predominantly forms sulfidic bonds, including mono-, di-, and polysulfidic bonds, that link ENR molecules. Additionally, chemical reactions among ENR molecules through the opened epoxirane ring products result in weaker -O-O- linkages (compared to sulfidic linkages) with a lower bond dissociation energy of approximately 184 kJ/mol [34]. In contrast, the bond association energy of the sulfur-sulfur bond or monosulfedic bond (-S-S-) is about 270 kJ/mol [35]. Therefore, under the high temperature and shearing conditions of the curing process, weaker chemical bonds such as -O-O- linkages are more likely to break, along with other weak polysulfide linkages (-S-Sx-S-), leading to the observed reversion curing curves. In Fig. 1(a), it is evident that an increase in the GP loadings in the ENR matrix results in a higher curing curve, indicating an increase in maximum torque and torque difference, as shown in Fig. 2. This indicates the reinforcing effect of GP particles in the ENR matrix due to their ability to interact with the polar functional groups present in ENR molecules. Similarly, the addition of CNT (3 phr) to ENR (ENR-CNT3) also raises the curing curve and maximum torque (Fig. 1(b)), along with the torque difference (Fig. 2) of the ENR nanocomposite. The reinforcement is further enhanced by adding the secondary filler GP to the ENR-CNT3 composite, with a loading range of 5 to 50 phr, forming an ENR-CNT3/GPx hybrid composite. This sudden increase in the curing curve and corresponding torque values can be attributed to the chemical interactions between the polar groups present on the surfaces of CNT and GP nanofillers and the opened epoxirane ring products in ENR molecules, as depicted in the proposed chemical interaction mechanism described in Fig. 3. These interactions also lead to an increase in bound rubber, crosslinking density, and stiffness of the composite, which will be further explained in the following sections.
The occurrence of chemical interactions can be confirmed by examining the representative FTIR spectra of different ENR nanocomposites, as illustrated in Fig. 4. The spectra clearly demonstrate a reduction in the intensity of the epoxide peak at a wavenumber of 870 cm− 1 [36] upon the addition of nanofillers. Additionally, the peak intensity of the hydroxyl groups (-OH stretching vibration in the range of 3200–3500 cm− 1), which are the products of opened ring epoxide groups, also decreases when fillers are incorporated. A notable observation is that the ENR-CNT3 and ENR-CNT3/GP20 hybrid composites exhibit a significant decline in the peak intensities of the OH groups compared to the ENR-GP20 composite and gum ENR vulcanizate. This indicates a higher level of chemical interaction between these functional groups and the polar functional groups present on the filler surface, as depicted in Fig. 3. Furthermore, the ENR-CNT3 and ENR-CNT3/GP20 composites show a higher peak intensity at a wavenumber of 1260 cm− 1 (indicating the C-O-C ether stretching vibration) [37] compared to the gum ENR vulcanizate (Fig. 4). This observation confirms the occurrence of a new form of -C-O-C- linkage resulting from the chemical interaction mechanism described in Fig. 3.
Figure 5 illustrates the scorch time (Ts2) and cure time(T90) of gum ENR and its compounds filled with graphite (ENR-GPx), carbon nanotubes (ENR-CNT3), and hybrid fillers (ENR-CNT3/GPx). It is evident from Fig. 5(a) that the incorporation of GP into the ENR vulcanizate to form ENR-GPx composites leads to a decrease in scorch time compared to the gum ENR compound. This reduction in scorch time can be attributed to the excellent heat-conducting properties of graphite, which facilitate the dissipation of heat generated during the curing process. This leads to an earlier initiation of the vulcanization reaction, resulting in a shorter time to the start of crosslinking reaction or scorch time. As the GP loading increases, heat conductivity also increases, leading to a greater reduction in scorch time. However, the cure time of the ENR-GPx composites increases with higher GP loading, as shown in Fig. 5(a). This can be attributed to the fact that graphite particles act as physical barriers, impeding the diffusion of curing agents and other additives into the ENR matrix. This hindrance slows down the curing process, leading to longer cure times. Additionally, the addition of graphite may alter the chemical interactions between the ENR matrix and the curing agents, thereby affecting the curing kinetics.
A previous study found that the critical percolation threshold concentration (PTC) for ENR-CNT composites was found to be 3 phr [38]. When this concentration is reached, interconnected CNT-CNT pathways form within the ENR matrix. These pathways facilitate the movement of electrons and phonons along the CNT networks embedded in the continuous ENR phase, thereby enhancing both electrical and thermal conductivity. Consequently, the vulcanization process is initiated earlier and the curing process is accelerated, resulting in a shorter scorch time compared to gum ENR compounds. The data in Fig. 5(b) clearly shows that the scorch time of ENR-CNT3 (at GP content = 0 phr) is shorter than that of gum ENR (at GP content = 0 phr in Fig. 5(a)). In contrast, Fig. 5(b) indicates that increasing the GP content in the ENR-CNT3/GPx nanocomposite does not have a significant effect on the scorch time. However, there is an increasing trend in the cure time of the ENR-CNT3/GPx hybrid composite with higher GP loadings. This can be attributed to the fact that as the GP loadings increase, the viscosity of the composite also increases. The higher viscosity makes it more challenging for the curing agent to penetrate and reach the ENR molecules, which are necessary to initiate and sustain the reaction between the curing agent and the rubber chains. Consequently, the curing process takes longer, resulting in the observed increase in cure time.
3.2 Payne effect
The Payne effect is a phenomenon that occurs in filled elastomers, where the size and concentration of filler particles dispersed within the elastomer matrix affect the mechanical properties of the composite material. Figure 6 illustrates the evaluation of the Payne effect in filled elastomer composites by measuring the storage shear modulus as a function of strain amplitude. The results clearly demonstrate that the addition of reinforcing filler to the gum ENR compound significantly increases the dynamic shear modulus. However, the storage moduli of filled ENR composites exhibit a slight decrease within the strain amplitude range of 1–10%, followed by a sharp decreasing trend within the strain range of 10–100%. This decrease in moduli with increasing strain amplitude is known as the Payne effect, which can be attributed to a partially reversible breakdown of the filler network [39].
The dynamic modulus of filled elastomer compounds is influenced by two types of properties: strain-dependent and strain-independent properties [40]. The strain-independent part is a combination of filler-rubber interactions, the contribution of the crosslinked rubber network, and the hydrodynamic effect of the filler. On the other hand, the strain-dependent part of the modulus arises from the filler-filler interaction, which leads to a non-linear decrease in the shear modulus as the strain amplitude increases. This effect can be primarily attributed to the Payne effect of the fillers dispersed within the rubber matrix. To quantitatively assess the Payne effect of different filled ENR compounds, the difference in storage moduli (ΔG') between the minimum (strain = 0.56%) and maximum (strain = 100%) strain amplitudes was compared, as shown in Table 3. The results indicate a significant difference in the Payne effect between ENR-GPx and ENR-CNT3 as well as ENR-CNT3/GPx hybrid nanocomposites. This difference may be attributed to the higher level of chemical interaction between CNT particles, as well as between CNT and GP particles, resulting in increased filler-filler interactions and a larger Payne effect. Furthermore, it is observed that the filler-filler interaction and the resulting Payne effect increase with increasing GP content in both ENR-GPx and ENR-CNT3/GPx hybrid nanocomposites. This is evidenced by a more pronounced breakdown of the filler networks, as indicated by a steeper gradient of the storage modulus-strain relationship in Fig. 6.
Table 3
Storage moduli (G') of gum NR, gum ENR, and their filled compounds with GP, CNT and CNT/GPx hybrid fillers at very low (G'0.56) and high (G'100) strain amplitudes, as well as their difference (ΔG').
Samples
|
G'0.56
|
G'100
|
ΔG' (G’0.56− G’100)
|
Gum ENR
|
70.71
|
51.22
|
19.49
|
ENR-GP10
|
85.33
|
55.94
|
29.39
|
ENR-GP30
|
89.42
|
48.48
|
40.94
|
ENR-GP50
|
98.75
|
45.01
|
53.74
|
ENR-CNT3
|
125.3
|
69.64
|
55.66
|
ENR-CNT3/GP10
|
129.6
|
57.58
|
72.02
|
ENR-CNT3/GP30
|
156.76
|
56.95
|
99.81
|
ENR-CNT3/GP50
|
198.05
|
55.94
|
142.11
|
3.3 Bound rubber (BdR) content.
Bound rubber (BdR) refers to the rubber portion in an uncured compound that cannot be extracted by a good solvent due to the adsorption of rubber molecules onto the filler surface. In Fig. 7, the total bound rubber content of ENR compounds filled with graphite, carbon nanotubes, and their hybrid nanofillers is illustrated. The results indicate that increasing filler loadings result in higher bound rubber content for both ENR-GPx composites and ENR-CNT3/GPx hybrid composites. However, the bound rubber content is higher in ENR-CNT3/GPx hybrid composites compared to ENR-GPx composites at corresponding graphite loadings. This can be attributed to the stronger filler-rubber interaction in ENR-CNT3/GPx composites, leading to the formation of a higher amount of bound rubber. This suggests increased physical adsorption and interaction, as well as chemisorption [41]. Nonetheless, some reports suggest that bound rubber primarily arises from a chemical phenomenon involving the interaction between the elastomer and the surface active filler particles [42, 43]. Therefore, the proposed chemical interaction mechanism between the polar functional groups of ENR molecules and nanofillers (i.e., CNTs and GP) shown in Fig. 3 is considered the main cause of bound rubber formation in the ENR nanocomposites. Additionally, the higher total bound rubber content in ENR-CNT3/GPx hybrid composites may be attributed to the possibility of stronger chemical interactions and bonding between ENR and the nanofiller surface, facilitated by the high polarity of CNT and GP surfaces. Figure 7 also reveals a strong correlation between the total bound rubber content and the storage modulus magnitude of the filled ENR composites in the low strain range of 1–10% (Fig. 5). In other words, the ENR-CNT3/GPx hybrid composites exhibit significantly higher storage shear moduli than ENR-GPx composites at a given GP loading. Furthermore, it is evident that the BdR follows the corresponding trend observed in the cure curves and their torque differences (Figs. 1 and 2, respectively).
3.4 Morphological properties
Stereo-microscope images and SEM micrographs were analyzed to visualize the formation of filler networks, as well as the dispersion and distribution of nanofillers in the ENR matrix, as the results are shown in Figs. 8 and 9, respectively. It is clearly seen that the gum ENR vulcanizate displayed a smooth fractured surface (Figs. 8(a) and 9(a)), with small white spots of ZnO particles. When 3 phr of CNT was added to the ENR matrix to form the ENR-CNT3 nanocomposite, different morphologies were observed (Figs. 8(b) and 9(b)). The SEM micrograph of the NR-CNT3 nanocomposite (Fig. 9(b)) revealed a fine dispersion of CNT networks with small CNT aggregates, indicating the formation of CNT networks within the ENR matrix. Furthermore, in the ENR-CNT3/GPx hybrid composites with different GP loadings at 10, 30, and 50 phr, distinct morphologies were observed in Figs. 8 and 9 ((c), (e), and (g)). That is, the ENR-GP10 composite (Figs. 8 (c) and 9(c)) exhibited larger GP aggregates compared to the CNT aggregates in the ENR-CNT3 composite (Figs. 8(b) and 9(b)). This can be attributed to the higher chemical interaction between the polar functional groups on the CNT surface and ENR molecules. Moreover, increasing GP loading in the ENR-GPx composites resulted in rougher fracture surfaces, as seen in Figs. 8 and 9 ((c), (e), and (g)). High-resolution SEM micrographs in Fig. 9((c), (e), and (g)) revealed an increasing size and number of GP agglomerations, indicating poorer dispersion and distribution of GP in the ENR matrix with increasing GP loadings. In contrast, the ENR-CNT3/GPx hybrid composites (Figs. 8 and 9((d), (f), and (h))) exhibited smoother surfaces compared to the ENR-GPx composite at a given GP loading. This indicates finer dispersion and distribution of the CNT/GP hybrid filler in the ENR matrix compared to GP in ENR-GPx composites. This can be attributed to the higher chemical interaction between the polar groups in ENR and the hybrid filler surface (Fig. 3), facilitating a more uniform dispersion and distribution of the CNT/GP hybrid filler within the ENR matrix. The dispersion and distribution of GP particles in the NR-CNT3/GPx hybrid composites were mainly attributed to the effective prevention of GP re-agglomeration by the CNT bundles, owing to the high interfacial energy between CNT and NR [14]. Therefore, the hybridization of CNT with GP can prevent the aggregation of carbon nanofillers and exhibit a synergistic effect in enhancing composite properties. This phenomenon is even more pronounced in polar rubber matrices like ENR, due to the synergistic influence of the chemical interaction among the polar groups of ENR and the nanofillers. This is also reflected in the greater torque differences observed for ENR-CNT3/GPx hybrid composites compared to their counterparts, ENR/CNT3 and ENR/GPx composites (Figs. 1 and 2). However, increasing GP loadings in ENR/GPx and ENR-CNT3/GPx resulted in larger GP agglomerations with poorer dispersion and distribution capability. This can be attributed to the formation of filler networks due to higher filler-filler interactions based on Van der Waals forces, dipole-dipole interactions, and hydrogen bonding among the filler particles [14, 44]. This can be well correlated to the increasing trend of the Payne effect (Fig. 5 and Table 3).
3.5 Tensile properties
The stress-strain behavior of a material is a fundamental mechanical property that characterizes its deformation under external stress. This study investigated the stress-strain behaviors of gum ENR and its compounds, which were filled with various fillers including CNT, GP, and a hybrid combination of CNT and GP with different GP loadings, as results shown in Fig. 10. It is clearly demonstrated that the Young's modulus, determined from the slope at the initial part of the stress-strain curve, increases as the GP loadings in both ENR-GPx and ENR-CN3/GPx composites increase. This increase is attributed to the reinforcement resulting from the chemical interaction between ENR molecules and carbon fillers that possess reactive functional groups, as depicted in Fig. 3. Furthermore, the trend observed in the Young's modulus aligns with the increasing trend of 100% moduli as the GP loadings increase, as shown in Fig. 11. Specifically, the 100% modulus of ENR-CNT3/GPx hybrid composites is higher compared to ENR-GPx composites. This can be attributed to the finer filler dispersion and distribution of GP in the ENR-CNT3/GPx hybrid composites, where CNTs effectively prevent the re-agglomeration of GP, as illustrated in Figs. 8 and 9. Consequently, the hybridization of CNTs with GP prevents the aggregation of carbon nanofillers, resulting in synergistically enhanced composite properties. This effect primarily arises from a larger filler surface that intimately interacts with the ENR matrix, facilitating the chemical interaction described in Fig. 3. Therefore, the mechanical properties are intricately linked to the morphological properties (Figs. 8 and 9), the total bound rubber content (Fig. 7), and the torque difference (Fig. 2), all of which exert a significant influence on the level of the chemical reaction and the resulting modulus (Fig. 11).
3.6 Electrical properties
The electrical conductivity of materials is a crucial property that characterizes their ability to conduct electricity under an applied electric field. The electrical conductivity can be used to verify the formation of filler pathways of GP-to-GP, CNT-to-CNT, and CNT-to-GP connections within the ENR matrix. These connections are observed in ENR-GPx, ENR-CNT3, and ENR-CNT3/GPx hybrid composites, respectively. Figure 12 displays the electrical conductivity of gum ENR and its composites filled with GP, CNT, and CNT/GP hybrid filler.
The results indicate that the inclusion of GP in ENR-GPx composites led to a slight increase in electrical conductivity as the GP loading increased compared to the gum ENR compound. This can be attributed to the incorporation of highly electrically conductive graphite (approximately 104 S/cm) [45], which forms a conductive network within the ENR matrix.
In contrast, the addition of 3 phr of CNT in ENR-CNT3 composites resulted in a significant increase in electrical conductivity. This abrupt increase can be attributed to the broader range of electrical conductivity exhibited by CNTs (approximately 103 to 105 S/cm) [46], along with the finer dispersion of CNTs within the ENR matrix as depicted in Figs. 8 and 9. This phenomenon promotes the formation of highly conductive pathways between carbon nanotubes (CNT-to-CNT pathway) rather than pathways between graphene platelets (GP-to-GP pathway). Figure 12 also demonstrates that incorporating different loadings of GP in ENR-CNT3 composites to form ENR-CNT3/GPx hybrid composites resulted in a substantial increase in electrical conductivity. Thus, the incorporation of GP enhances the dispersion and distribution of CNT and GP fillers, leading to the formation of interconnected infinite networks or end-to-end connections of the carbon fillers at the percolation threshold concentration (ϕc). This relationship between electrical conductivity and filler content is illustrated in Fig. 13.
In Fig. 13, a significant increase in electrical conductivity is observed for ENR-CNT3/GPx hybrid composites compared to ENR-CNT3 and ENR-GPx composites, especially when the GP content in the hybrid composite is increased. This compelling evidence confirms the synergistic effect of the CNT-GP hybrid filler when incorporated into the ENR matrix. In contrast, the introduction of graphene platelets in ENR-GPx composites only results in a marginal improvement in electrical conductivity as the GP content increases. It is worth noting that the maximum achieved electrical conductivity of ENR-GPx composites is considerably lower than that of both ENR-CNT3 and ENR-CNT3/GPx nanocomposites. This observation can likely be attributed to the inherently lower electrical conductivity of GP particles, coupled with their tendency to form larger agglomerations within the ENR matrix, as depicted in Figs. 8 and 9. On the other hand, the electrical conductivity of ENR-CNT3 composite suggests that the primary factor influencing conductivity enhancement in ENR composites is the loading of CNTs. Figure 13 also demonstrates that the electrical conductivity of ENR-CNT3/GPx hybrid composites continuously increases as the GP content increases. Moreover, higher conductivities are observed throughout the range of incorporated GP compared to ENR-CNT3 composites. Furthermore, the addition of graphene platelets in the range of 5–15 phr within the ENR-CNT3/GPx composites appears to contribute to an increase in electrical conductivity. This can be attributed to the formation of partially connected filler networks comprising both CNTs and GPs, resulting in the creation of a "dead-arm" structure or end-to-end connection, as described in previous work [14]. This range of GP loadings facilitates electron transport throughout the ENR matrix through the tunneling effect. It should be noted that the tunneling effect is influenced by various factors, including the interphase layer surrounding nanoparticles, the thickness of the bound rubber layer, filler dimensions, filler conductivity, interphase thickness, waviness, fraction, and content of filler network, as well as tunneling distance [47, 48]. Therefore, in ENR-CNT3/GPx composites (Fig. 13), increasing the GP loading further affects the formation of a complete connected network of conductive fillers in the ENR matrix, resulting in maximum conductivity at a GP loading of around 20 phr. This enables the formation of maximum CNT-GP-CNT connections, which promote the tunneling effect of electrons in the ENR nanocomposites.
Furthermore, the trend of electrical conductivity depicted in Figs. 12 and 13 consistently aligns with the dielectric properties shown in Fig. 14. It is important to note that the dielectric constant or relative permittivity of a material reflects its ability to store electrical energy in the presence of an electric field. Figure 14 clearly illustrates that ENR-CNT3 and ENR-CNT3/GPx composites exhibit significantly higher dielectric constants compared to ENR-GPx composites. Moreover, the incremental increase in dielectric constant with increasing GP loadings within both composites is relatively low. This behavior can be attributed to several factors, including the formation of a highly interconnected network of CNTs, the conductivity enhancement provided by the graphite particles, and the inherent elastic properties of the elastomer. Therefore, it can be concluded that the ENR-CNT3/GPx composites possess high conductivity and dielectric constant, rendering them suitable for various applications in electronic devices, electrodes, actuators, and sensors. These composites can sustain electric currents while retaining superior mechanical and elastic properties due to the formation of networks by CNT bundles and GP platelets, which reduces the accumulation of electron charges at the ends of filler particles [14]. Consequently, electrons can flow more easily throughout the ENR matrix, significantly enhancing the curing properties (Fig. 1), electrical conductivity (Figs. 12 and 13), and dielectric constant (Fig. 14) of the ENR composites.