Introduction

Polymer nanocomposites (PNCs) have become one of the most significant subjects of research since from 1984, when the Toyota Research Group of Japan reported the development of PA6–clay nanocomposites for the first time [1]. The high aspect ratio, large surface area and substantial reinforcement of nanoparticles compared to their conventional counterparts made researchers to put great efforts on the development, characterization and modeling of PNCs [2, 3]. However, the key issue in manufacturing nanocomposites is to obtain a complete dispersion of filler in the matrix and maximum interfacial compatibility. For 2D nano sheets, the individual layers have much higher aspect ratio than typical microscopic aggregates, which have to be exfoliated to yield high performance nanocomposites [4]. Nanoclays impart outstanding mechanical and permeability properties to elastomer nanocomposites due to their large interfacial area and attractive interactions per unit volume with the matrix [2, 5, 6]. It is reported that the dispersion state of the organoclay in a rubber matrix is mainly influenced by the nature of aggregated platelets [7–9], compounding conditions [10, 11] (shear rate and temperature) and polarity of the rubber matrix [12, 13]. Rubbers exhibit high melt viscosities during melt mixing because of their high molecular weight and this paves the path for the generation of high shear stresses for shearing and peeling apart of the silicate layers [14]. In addition to melt compounding, several other techniques like in situ polymerization [15, 16], solvent-assisted mixing [17], and emulsion compounding, are also used to disperse fillers in rubber matrices at a nano-scopic scale.

Apart from the clay nanocomposites, great interest has been laid on rubber/graphene nanocomposites to get improvement in mechanical, electrical and permeability properties [15, 17–20]. Graphene is an atomically thick two dimensional sheet of sp2-bonded carbon atoms (2-D) arranged like a honeycomb structure with ultra-high specific surface area (theoretical limit ~2,630 m2/g) [21]. Moreover, outstanding mechanical, electrical, thermal and gas impermeability properties were also demonstrated [22–24] for this material. At low concentrations, barrier properties of graphene sheets are ~25–130 times higher than that of clay nano fillers [25] and more effective (by a factor of 1 order of magnitude or higher) than predicted by the modified Nielsen, Cussler and Bharadwaj theories [26–28]. However, to fulfill these expectations, the fillers have to be uniformly dispersed in the elastomer matrix without any agglomeration. Huiqin et al. [17] reported well-dispersed modified graphene (MG) in IIR (poly(isobutylene-co-isoprene)) composites through solution mixing process, with dramatically enhanced mechanical properties. Yanhu et al. [15] used an ultrasonically assisted latex mixing and in situ reduction (ULMR) process to prepare natural rubber latex (NRL)/graphene composites. The effect of functionalized graphene on natural rubber (NR) and styrene butadiene rubber (SBR) matrices are demonstrated by Prud'homme et al. [19]. Xin et al. [20] explained the reinforcement of hydrogenated carboxylated nitrile–butadiene rubber (HXNBR) with exfoliated graphene oxide (GO). The strong interfacial interactions between the oxygen-containing functional groups on the surfaces of GO nano sheets and the carboxyl groups in HXNBR improved the mechanical properties, even at very low level of GO loading.

IIR is a copolymer of isobutylene (97–98 %) and a small amount of isoprene (2–3 %). Since its commercialization in 1943, the rubber finds enormous applications in many areas because of its excellent properties such as impermeability/air retention, weathering resistance, ozone resistance, vibration dampening, and sealing efficiency. It has good flexing properties as well, resulting from the low levels of unsaturation between long polyisobutylene segments [29–32]. This synthetic rubber is used in a wide range of applications such as inner tubes, tyre liners, curing bladders, air springs, sealants, gas pipe coating and gaskets, air conditioner hoses, cable insulations, jacketing, roof membranes and sporting goods [29, 33, 34]. However, various nano fillers and other additives are added to IIR to improve its performances further. IIR/clay nanocomposites are difficult to prepare because of the hydrophobicity of IIR and its poor interaction with clay silicate layers [35]. However, both solution and melt mixed samples having similar intercalated structures are reported, [36–40] with better mechanical and gas-barrier properties for the former. This is due to the fact that during curing of solution mixed samples under high pressure the solvent molecules within the silicate inter layers favor the motion of the intercalated IIR chains [41]. By reinforcing IIR with swollen clay platelets, Liang et al. introduced a new method of composite fabrication [42]. Samadi and Kashani [43] observed cloisite10A as the most efficient reinforcing clay for butyl-rubber compounds and the same is used in this work. Maiti et al. [44] succeeded in dispersing filler in rubber via radical grafting reaction [44]. Grafting of IIR with maleic anhydride (MA) can introduce polar groups on the non polar IIR backbone, which improves the IIR/organoclay compatibility depending on the grafting level [45, 46]. Makoto et al. [35] performed solvent assisted MA grafting on IIR, whereas Gunberg and Ridgewood [47] employed a solvent-free method.

Up to now, many researchers worked on IIR/clay nanocomposites, but none of them could achieve much improvement in the mechanical and barrier properties simultaneously. In the present work, for the first time, we compare the interfaces of IIR compounds filled with the two widely used two dimensional fillers, cloisite10A and thermally reduced graphene (TRG). Secondly, the influence of MA grafting on the cloisite 10A–rubber interface is evaluated by carefully comparing the samples with different MA-grafted poly(isobutylene-co-isoprene) (MA-g-IIR) content. MA-g-IIR was used along with IIR to improve the dispersion of clays in the prepared composites. Morphological evidence was obtained from FTIR, SEM, AFM, XRD and TEM techniques and correlated with the filler and elastomer interactions existing at the composite interfaces. Finally the mechanical and gas permeability properties of clay and graphene composites were compared and both properties are found to be respectively depending on the filler–filler and filler–elastomer interactions in the composites.

Experimental procedure

Materials

IIR was purchased from Exxon Company and the nanoclay, Cloisite10A (density 1.90 g/cm3) was obtained from Southern Clay Products. Natural flake graphite, MA, curing agents, benzoyl peroxide, xylene and all other minor reagents were purchased from Aldrich, France.

Preparation of TRG from GO

Graphene is prepared from graphite by following chemical approach. This technique involves two key steps: (1) oxidation of graphite to GO and (2) thermal reduction of the prepared GO to TRG. Firstly, GO is prepared by the so-called improved GO synthesis method [48]. For the reduction of GO, the thermal method was preferred rather than chemical method [49, 50] so that better dispersion is achieved with TRG. For thermal reduction, nitrogen gas was passed through GO for 30 min in sealed container to create an inert atmosphere on GO surface and the resulting product was placed in pre-heated muffle furnace at 1,000 °C for 30 s.

Preparation of MA-g-IIR

MA-g-IIR was prepared according to the method described by Makoto et al. [35] method. First, 100 g of IIR and 42 g of MA were dissolved in 400 ml xylene at 110 °C. Benzoyl peroxide (1.86 g) was dissolved in 100 ml xylene and delivered drop wise into the IIR solution over a period of 10 min. The IIR solution was stirred at 110 °C for 3 h, and then poured into acetone, resulting in the precipitation of MA-g-IIR. The MA-g-IIR was dried at 80 °C in a vacuum oven till it attained a constant weight. Using the size exclusion chromatographic technique the molecular weights of IIR and MA-g-IIR were determined and is given in Table 1. Grafting is confirmed on the basis of the decrease in molecular weight observed for MA-g-IIR.

Table 1 Molecular weight of poly(isobutylene-co-isoprene) and maleic anhydride grafted poly(isobutylene-co-isoprene) from size-exclusion chromatography (SEC)
Full size table

Preparation of nanocomposites

Solution mixing process was employed for synthesizing IIR nanocomposites in order to get better dispersion cloisite10A and graphene in IIR. A constant filler loading of 5 phr is used for all samples. At first, filler was sonicated in THF for 1 h and then mixed with IIR and or MA-g-IIR/THF solution using mechanical stirrer for 3 h. All composites were dried at 80 °C in a vacuum oven till it attained a constant weight. The cured samples were obtained by mixing the nanocomposites with the curing agents such as stearic acid at 2 phr, zinc oxide (Activator) at 5 phr, 2-bisbenzothiazole-2, 2′-disulfide (MBTS) at 0.5 phr, tetramethyl thiuram disulfide (TMTD) at 1 phr and sulfur (elemental sulfur S8) at 1.5 phr, in an internal brabender mixer at 20 rpm and at 50 °C for 30 min. The samples were cured by press molding at 160 °C for their optimum cure times determined from the Moving Die Rheometer. The composition of the prepared samples is listed in Table 2.

Table 2 Data of compounding, poly(isobutylene-co-isoprene) (IIR), maleic anhydride grafted poly(isobutylene-co-isoprene) (MA-g-IIR)
Full size table

Characterization

The mechanical measurements were carried out in an INSTRON 5566A universal tensile machine at a test speed of 10 mm/min. Dog bone-shaped samples were used for the tests and the measurements were done at room temperature (25 °C). The stress relaxation was monitored for a constant strain of 40 % for all the measurements for 2 h with 100 mm/min rate to attain the required strain. XRD measurements were carried out in an expert model of Philips diffractometer with Cu Kά radiation generator (1.5404 Å, 40 kV, 40 mA). The ultra-thin transmission electron microscopy (TEM) samples of 70-nm thickness were cut using a cryogenic ultramicrotome Leica ultracut UCT at −90 °C and images were taken using a JEOL JEM-1400 electron microscope at 100 kV. Scanning electron microscopic images of the cryo cut samples were taken with JEOL JSM-6460LV SEM. Atomic force microscopy (AFM) images were recorded with diCaliber Veeco Instrument. In order to measure the molecular weights, a Shimadzu SP 10Avp UV dual wavelength detector (λ 1 = 254 nm and λ2 = 280 nm) was used with THF solvent at room temperature. The oxygen transmission rate (OTR) across the material is monitored using an oxysence 400B device. Payne effect was observed from the strain sweep experiments from 0.01 % to 10 % at a frequency of 0.5 Hz using Metravib dynamic mechanical analyser.

Results and discussions

FTIR analysis and the grafting degree

It is established that the increase in the grafting density for polymers results in decrease in the molecular weight [37]. The rate of chain scission depends on concentration of benzoyl peroxide, temperature, reaction time, etc. The decrease in molecular weight observed for MA-g-IIR from the GPC data (Table 1) is because of the free radical induced chain scission of IIR by benzoyl peroxide during grafting. In addition to the GPC technique, FTIR spectroscopy gives confirmation for the MA grafting on IIR.

Figure 1 shows the FTIR spectra of raw IIR and MA-g-IIR. Contrasted to the FTIR of raw IIR, the appearance of new absorbance peaks at 1,860 and 1,785 cm−1 (corresponding to two V C=O vibrational stretching modes, symmetric and asymmetric in MA) in MA-g-IIR indicate the MA grafting onto the IIR backbone. FTIR is also used for determining the grafting degree following Tosaka et al. [51] method. The peaks located at 1,780 and 2,728 cm−1 (Fig. 1) can be assigned to the stretched vibration absorbance of the carbonyl and methyl groups, respectively. Since the methyl groups on the IIR were unaffected during grafting, the ratio of the absorbance at 1,780 cm−1 over that at 2,728 cm−1 (A 1,780/A 2,728) was taken as a measure of the relative grafting degree. It is observed that the level of grafting depends on the experimental conditions (temperature, MA content, etc.). The grafting time was determined to be 3 h since this yields acceptable grafting level and chain scission degrees. The grafting density of 0.75 % is found as the best and was chosen for the future experiments.

Fig. 1
figure 1

FT-IR spectra of a neat poly(isobutylene-co-isoprene) (IIR) and b maleic anhydrate grafted poly(isobutylene-co-isoprene) (MA-g-IIR)

Full size image

Morphology of graphene and nanocomposites

The thickness of GO was determined by AFM and after thermal reduction, graphene sheets were characterized by XRD. Figure 2 demonstrates the AFM images which indicate the thicknesses of the exfoliated GO layers to be ~1.5 nm while width and length are in the range of ~1 μm. The XRD patterns of natural graphite (NG), IIR, IIR/TRG (RG) (Table 2) nanocomposites are presented in Fig. 3a. The XRD of TRG shows no characteristic peak, indicating its complete exfoliation. The appearance of a broad peak at 15° in Fig. 3a, for IIR and IIR/RG films, is due to the amorphous phase of IIR [17]. The diffraction ascribed to NG (2θ = 26.5°) does not appear in the XRD pattern of the IIR/TRG composite, indicating the complete exfoliation of TRG in the IIR matrix.

Fig. 2
figure 2

AFM images of graphene oxide platelets deposited on a smooth silica surface

Full size image
Fig. 3
figure 3

WAXD patterns of a thermally reduced graphene (TRG), IIR, IIR/TRG and natural graphite (NG); b IIR/cloisite10A nanocomposites containing various amounts of MA-g-IIR

Full size image

Figure 3b shows the XRD profiles of cloisite10A, RC, M2, M4, M6, M8 and MC nanocomposites. The (001) diffraction of cloisite10A at 2θ = 4.8°, corresponds to an inter-layer spacing of 1.83 nm. For the composite samples RC, M2 and M4, the characteristic (001) peak of the cloisite10A is shifted to 2θ = 1.92°, corresponding to a basal spacing of 4.61 nm. The position of the (001) peak basically does not change with the improvement of MA-g-IIR content in the nanocomposite, while the peak intensity decreases significantly and peak broadness increases. M6, M8 and MC, exhibit no characteristic peak in the XRD trace because of the complete loss of regularity between the clay layers. Owing to the low molecular weight, the MA-g-IIR can easily penetrate into the agglomerated nanoplatelets, strongly interact with the platelets’ edges and can start the peeling off process that yields exfoliation [52]. However, it is difficult for XRD to draw definitive conclusions about the dispersion of clay in nanocomposites. Thus, microscopic techniques such as SEM and TEM are necessary to provide an actual image of the clay layers and to permit the identification of the nanocomposites morphology. Since the resolution of SEM is not enough to picture the dispersed platelets, TEM micrographs for the RC and M6 nanocomposites are taken as given in Fig. 4.

Fig. 4
figure 4

TEM photographs of a poly(isobutylene-co-isoprene) (100 phr)/(5 phr) cloisite10A, (RC) and b poly(isobutylene-co-isoprene) (40 phr) + maleic anhydride grafted poly(isobutylene-co-isoprene) (60 phr)/(5 phr) cloisite10A, (M6)

Full size image

Since clay has much higher electron density than neat polymers, it appears dark in TEM images. TEM micrograph for IIR/cloisite10A (RC) nanocomposites shows that the clay layers consist of highly multilayered stacks of about 20–100 nm (Fig. 4a). On the other hand, in the micrographs of IIR/MA-g-IIR/cloisite10A (M6) nanocomposites, there is better dispersion of clay in the MA-g-IIR/IIR matrix (Fig. 4b). This is evidenced from the disordered single platelets observed in TEM image (Fig. 4b) which confirms the existence of a mixed intercalated/delaminated structure. The homogeneous dispersion of TRGs (Fig. 5a) in IIR is well clear from the SEM micrograph for IIR/TRG (RG) nanocomposites (Fig. 5b) itself. The floating particles (about ~0.4 μm in size) observed in the image is that of zinc oxide.

Fig. 5
figure 5

SEM photographs of a thermally reduced graphene and b poly(isobutylene-co-isoprene) (100 phr) + thermally reduced graphene (5 phr)

Full size image

Viscoelastic behavior

Understanding the viscoelastic properties of elastomer nanocomposite is of great importance to get a fundamental knowledge of the composite microstructure. The inter-particle and elastomer–filler interactions can strongly influence both linear and nonlinear viscoelastic responses [53]. Rheology is therefore an effective tool for quantifying nano filler dispersion in an uncured molten matrix. The effect of organoclay and MA-g-IIR on the apparent melt elastic modulus, G’ (Pa) value of IIR is shown in Fig. 6. Comparison of the linear viscoelastic responses shows the significant effect of the clay, particularly at low frequencies (0.1–0.03 Hz). Shear rate also influences the composite morphology to a great extent. At low shear rates, the silicate platelets are well separated causing the viscosity of the melt to increase strongly, whereas shear thinning behavior is noticed at higher shear rates due to the possible orientation effect of the fillers [54].

Fig. 6
figure 6

Variation in elastic modulus, G′ for neat IIR, IIR/cloisite10A, IIR/graphene MA-g-IIR/cloisite10A and IIR/cloisite10A nanocomposites containing various loadings of MA-g-IIR at a test strain of 1 % and at a temperature of 150 °C

Full size image

It is observed that, at low frequencies M6, M8 and MC composites give higher G′ values whereas RC behaves identically to the neat IIR matrix. M2 and M4 samples having low MA-g-IIR content also do not show significant change from the behavior of RC. The G′ increase observed at low frequencies arises from the good dispersion of the platelets that increase the cohesion of the molten nanocomposites. Chemical bonds created between the MA rings of MA-g-IIR and the OH groups on the edges of the clay platelets were claimed to be at the origin of a peeling off of the clay platelets from the stacks [52]. This is consistent with the conclusion drawn by Li et al. [55] that the intensive interaction between the exfoliated silicate layers and polymer chains increases the complex viscosity and causes a marked shear thinning at low frequency values. This is pictured as well-structured M6, M8 and MC nanocomposites revealing the good dispersion of the platelets in the matrix. In the case of IIR/TRG (RG) also, the G′ is much higher than the neat IIR; however this increase is not as much as observed in the case of M6. This is due to the existence of both filler–filler and filler–elastomer interactions in M6 and the presence of only the latter in RG.

Mechanical properties

The mechanical properties of IIR, RC (IIR/cloisite10A), M2, M4, M6, M8, MC and RG nanocomposites, quantified by tensile stress–strain measurements are summarized in Table 3 and Fig. 7. Grafting of IIR with MA improves mechanical properties of the nanocomposites due to the obtained better dispersion of cloisite10A in the matrix [45, 46] and thus the strong interface. Partial exfoliation of nano clay, depending on the grafting level was clear from TEM images. Makoto et al. [35] noticed an improvement in the barrier properties without any significant improvement in mechanical properties for the organo modified clay/MA-g-IIR nanocomposites (with very high grafting 1.75 %). In this study, we prepared a very low grafted system 0.75 %, by reducing the reaction time and simultaneously achieved significant improvement in both mechanical and permeability properties as discussed hereafter. In addition to the tensile strength, the elongation at break value is also enhanced. More clearly IIR/cloisite10A shows ~0.15 times improvement in tensile modulus compared to the unfilled matrix and it increases with the amount of MA-g-IIR up to 60 phr and then started decreasing. For tensile strength and elongation at break of M6, the increase is ~4.1 and ~2 times higher than that of the unfilled IIR. It can be observed that the dispersed nanoclay layers effectively improve the mechanical properties (strain and elongation at break) of M6 better than other composites. M6 contains sufficient amount of MA-g-IIR chains to disperse the clay and the remaining ungrafted IIR provide additional strength. Because of the higher molecular weight compared to M8 and the better dispersion of clay compared to M4, M2 and RC, the M6 is found to be the strongest IIR clay nanocomposite.

Table 3 Mechanical and Oxygen gas permeability properties of neat IIR, IIR/TRG, IIR/cloisite10A, MA-g-IIR/cloisite10A and IIR/cloisite10A nanocomposites containing various loadings of MA-g-IIR
Full size table
Fig. 7
figure 7

Young’s modulus (MPa), tensile strength (MPa) and elongation at break (%) of neat IIR, IIR/graphene, IIR/cloisite10A, MA-g-IIR/cloisite10A and IIR/cloisite10A nanocomposites containing various loadings of MA-g-IIR

Full size image

The tensile modulus (Young’s modulus) of IIR/graphene is ~2.4 times higher than that of the unfilled matrix because of the strong filler–rubber interactions in RG than all other IIR/clay nanocomposites. The high specific surface area and the two-dimensional geometry of TRG result in improved mechanical interlocking and adhesion with polymeric chains [56, 57]. Xin et al. [20] reported that the GO/HXNBR composite with 0.44 vol.% of GO has a balanced mechanical properties whereas at 1.3 vol.% loading, they observed a decrease in tensile strength and an increase in Young’s modulus for the GO/HXNBR composite compared to the neat HXNBR. The same conclusion is drawn for our RG system for which 5 phr level of graphene loading yields the highest modulus but not the highest strain and elongation at break when compared to clay nanocomposites which may be due to the lack of strong filler–filler interactions.

Stress relaxation

Dynamic mechanical tests can better understand the nanocomposites properties. The process of stress relaxation takes place due to both physical and chemical changes in the rubber matrix. The initial (reversible) change is caused by relaxation of the polymer chains and fillers. Long-term (irreversible) changes occur due to chemical reactions over time. A partial explanation of this can be provided by the theory of “strain amplification” [58] which assume greater strain in the polymer phase of a filled matrix than the overall strain due to the inextensibility of the filler. Stress relaxation rates, in general, increase with increasing polymer extension, thus filled rubbers would be expected to show higher relaxation rates [59, 60]. Other reasons for the higher relaxation rates in filled rubbers may, in some cases, are associated with filler–filler or filler–polymer interactions breakdown. The stress relaxation of polymers is attributed to chain motion and orientation, disentangling of polymer chain network strands, and deformation and rupture of micro domains and crosslinks [61]. If stress relaxation occurs, part of the energy stored in the material will be dissipated and part of deformed chains cannot retract. In the case of filled rubbers, desorption of polymer chains from the filler surface occur and the dewetting starts or propagates during the time of observation and the rate increases greatly. It was already reported that the extent of such breakdown varies with the nature of the filler [62]. To explain the observed behavior, the experimental relaxation curves of the composites were fitted to the stretched exponential Kohlrausch equation (Eq. 1) given by

$$ \frac{{{\sigma_{\mathrm{t}}}}}{{{\sigma_0}}}=\frac{{{\sigma_{\infty }}}}{{{\sigma_0}}}+\frac{{{\sigma_1}}}{{{\sigma_0}}}\exp \left[ {-{{{\left( {t/\tau } \right)}}^{\beta }}} \right] $$
(1)

where σ ∞/σ 0, σ t/σ 0, τ, and β are the fitting parameters. Here, τ is the relaxation time, β is the stretching parameter (0 < β ≤ 1), σ ∞/σ 0 is the transient stress, σ t/σ 0 is the limiting stress and \( \left[ {\left( {{\sigma_0}-{\sigma_{\infty }}} \right)/{\sigma_0}} \right]\times 100 \) is the relaxation ratio. The value of σ/σ 0 is called the normalized stress and the dependence of this on time is plotted as the stress relaxation curve. The best fit for the experimental values with the parameters of Eq. 1 are given in Table 4. From the fitting parameters, the relaxation time τ, and the stretching exponent β were estimated in order to understand the mechanism of the relaxation processes in the filled rubber composites.

Table 4 Fitting parameters of Eq. 1
Full size table

Figure 8 shows the stress relaxation curves [(σ t/σ 0) vs. time, t] for the neat IIR, IIR/TRG, IIR/cloisite10A, MA-g-IIR/cloisite10A and IIR/cloisite10A nanocomposites containing various loading of MA-g-IIR at a constant elongation of 40 %. It is observed that the stress decays with time until it approaches an equilibrium value. To evaluate the effect of the filler–polymer interactions, the values for IIR/TRG (RG) and IIR/cloisite10A (RC) are compared. Since the reinforcing mechanisms are different for the cloisite10A and graphene filled systems, their stress relaxation behavior can be originated from different sources. For RC, the relaxation in stress can be mostly attributed to the rupture of cloisite10A–IIR interactions, whereas it is not enough for the interaction breakage in graphene–IIR system. The addition of MA-g-IIR to IIR/cloisite10A (RC) decreases the overall chain flexibility and thus leads to a slow relaxation process. The results indicate that the addition of MA-g-IIR has a very strong effect on the crosslink density of the composites due to its influence on the physical and chemical interactions of IIR and cloisite10A (in RC). According to Chih-Cheng et al. [63], the stability of rubber–rubber interaction is stronger than filler–rubber and filler–filler interactions due to the hydrogen bonds present (Fig. 8b) [64]. The increase in the MA-g-IIR content led to the increase in the relaxation time (Table 4), which is again due to the strong filler–rubber interactions. This higher degree of polymer–filler interactions is further evidenced from the higher contribution in the progressive failure in filler–rubber bonding. For IIR/TRG (RG), the stress relaxation time is lower than that of MA-g-IIR/cloisite10A (MC), even though they have good reinforcement (from Young’s moduli) with rubber, but MC has strong filler–filler interaction than RG. The strong filler–rubber and filler–filler interactions are the reasons for the greater responses in the stress relaxation experiments in MC.

Fig. 8
figure 8

a Stress relaxation curves, time dependence on the normalized stress for neat IIR, IIR/graphene, IIR/cloisite10A, MA-g-IIR/cloisite10A and IIR/cloisite10A nanocomposites containing various loading of MA-g-IIR. (The dotted lines represent the curve fits.) b Stability of bonds responsible [63]

Full size image

Payne effect

The dependence of strain on the storage moduli of nanocomposites at 25 °C is studied (Fig. 9), and the Payne effect is observed. The filled composites show obviously higher storage modulus (G′) than neat rubber [65, 66]. However, the rate of MA grafting has a strong positive influence on the Payne effect due to the significant reinforcement. Since the storage modulus is related to the crosslink density for the filled composites (Eq. 2), the interactions within the composites can be deduced from storage moduli values.

Fig. 9
figure 9

Strain dependence of the storage modulus at 25 °C temperature for neat IIR, IIR/graphene, IIR/cloisite10A, MA-g-IIR/cloisite10A and IIR/cloisite10A nanocomposites containing various loading of MA-g-IIR (dotted lines represent the curve fits)

Full size image
$$ G\prime =N{k_{\mathrm{B}}}T $$
(2)

where k B is the Boltzmann constant, T is the temperature and N (N total = N c + N st + N i) is the crosslink density of the filled network. N c is the chemical crosslink density and bonds from entanglement, N st is the number of rubber–filler stable bonds per unit volume and N i is the number of rubber–filler unstable bonds per volume unit of the material.

The filler–filler and filler–rubber interactions based on Payne effect (decrease in G′ with increase in strain amplitude) can be explained by several models. The well-known Maier and Göritz [64] model mainly considers filler–rubber interactions with stable (strong) and unstable (weak) filler–rubber bonds (Eq. 3). The present composite systems studied here were observed to be in good agreement with this model (Fig. 9) revealing the significance of filler–rubber interactions.

$$ G\prime \left( \gamma \right)=G_{\mathrm{st}}^{\prime }+G_{\mathrm{i}}^{\prime}\left( {\frac{1}{{1+c\gamma }}} \right) $$
(3)

where G st′ = (N g + N st)k B T and G i ′ = N i k B T. G st′ is the value of G′ at high deformations and G I′ represents the Payne amplitude.

The number of rubber–filler stable bonds per unit volume for the neat IIR and various IIR nanocomposites are calculated based on this equation and given in Table 5. The elastic modulus of the composite comes from two factors, the pure rubber contribution and the filler contribution based on the filler/rubber interface. The number of rubber–filler stable bonds per unit volume for neat IIR (N st = 0.12 × 1026/cm3) is lower than all its nanocomposites due to the lack of additional cross links from the filler side. IIR/graphene (RG) shows the highest number of stable bonds (N st = 3.65 × 1026/cm3) among all nanocomposites because of the strong interactions between rubber and graphene. Among the clay nanocomposites, MC has highest number of stable bonds. Figure 9 also shows the increase in the number of stable bonds with the amount of MA-g-IIR. This is attributed to the increased specific surface area with the increase in MA-g-IIR content, which can improve the number of bonds and thus the filler–rubber bond formation. The same conclusion can be drawn from the analysis of the unstable bonds as well.

Table 5 Fitting parameters of Eq. 3
Full size table

Oxygen transmission rate

One of the important properties for IIR is its low gas permeability. The incorporation of layered structure nano fillers into polymer matrices has shown to be highly effective in reducing gas and solvent permeability [30–32, 42, 67]. The OTR values of IIR, RG, RC, MC, M2, M4, M6, and M8 are included in Table 3. A lower OTR is detected for RG when compared to the clay reinforced samples. By following the simple diffusion experiments using toluene and by calculating the diffusion coefficient values, the aspect ratios of the two fillers clay and graphene were estimated to be 108 ± 8 and 130 ± 2, respectively. The relatively higher aspect ratio of graphene in IIR matrix compared to clay platelets is one of the main reasons for the higher OTR rate in the clay sample. Factors like permeate interaction with filler and filler–polymer interactions [29–32],also affect OTR values. It has been proposed by Hailin et al. [68] that due to the poor compatibility of the silica surface and the polymer, the polymer chains could not tightly contact the silica nanoparticles, and thus it forms a narrow gap surrounding the silica particles. Indeed, graphene has more surface area (~2,630 m2/g) than clay (~750 m2/g) and, gas molecules need more time to cross graphene barrier than clay barrier.

Long tortuous path and high width are evidenced for IIR/TRG composite compared to the grafted and ungrafted IIR/cloisite10A composites because of the high surface area of TRGs and its better exfoliation in the elastomer medium. Well-exfoliated graphene sheets are evident from the XRD and TEM analysis. In fact the larger aspect ratio of graphene sheets (130 ± 2) over the clay (108 ± 8) leads to higher surface area and thus the better dispersion of graphene in IIR. All these point out towards the improved barrier property of RG. Among the cloisite10A composites, MC exhibit lower gas permeability than all others because of nanometer level dispersion of cloisite10A. In between M8 and M6, the former exhibit the same level of gas permeability of MC. But in the case of M6, the well dispersion (from XRD and TEM) and strong interface causes difficulty for the gas molecules to pass through the tightly crosslinked system [68, 69] compared to RC, M2 and M4, thus making MC a mechanically strong impermeable membrane. The orientation of platelets strongly depends on the barrier properties of composites as explained by Bharadwaj model [26]. But here, the same level of orientation is assumed for all composites due to the exactly similar preparation techniques and only the nature of fillers and filler polymer interactions are addressed in order to explain the permeability.

Even though the properties of clay and graphene filler reinforced composites are compared in this work, both these fillers are different in their aspect ratios, nature and interactions with the polymer chains. The correlation of molecular interactions occurring inside the composites and the more focused permeability application makes this study very meaningful.

Conclusions

The role of fillers like graphene and cloisite10A as well as the influence of grafting was studied in the IIR rubber matrix for a better understanding of the nano filler dispersion, interactions, mechanism of reinforcement and oxygen diffusion. We highlighted the possibility of tuning the grafting level to obtain MA-g-IIR/IIR ratio that maximizes rubber–filler interactions and act as physical crosslinking for the 60 phr MA-g-IIR/cloisite10A system. Very good level of filler dispersion is observed at 5 phr level, which is maintained throughout the study. Young’s modulus, Stress relaxation and Payne effect values are found to be enhanced for MA-g-IIR/IIR/cloisite10A composites. Due to the high aspect ratio and good filler–polymer interactions, the graphene nanocomposites exhibit a good enhancement in Young’s, storage modulus and decrease in oxygen permeability. The remarkably improved impermeability of IIR/graphene nanocomposites compared with the optimized IIR/cloisite10A nanocomposites arises also from the high specific surface area of graphene that hinder oxygen diffusion. Thus, the composite properties are successfully correlated with the filler elastomer molecular interactions taking place in IIR composites depending on the degree of grafting and nature of filler as well as the composition of composites. In the future experiments we are trying to find the synergistic effect of clay and graphene to improve the mechanical and permeability of rubber simultaneously.