Properties of Silicone Rubber-Based Composites Reinforced with Few-Layer Graphene and Iron Oxide or Titanium Dioxide

The increasing demand for polymer composites with novel or improved properties requires novel fillers. To meet the challenges posed, nanofillers such as graphene, carbon nanotubes, and titanium dioxide (TiO2) have been used. In the present work, few-layer graphene (FLG) and iron oxide (Fe3O4) or TiO2 were used as fillers in a room-temperature-vulcanized (RTV) silicone rubber (SR) matrix. Composites were prepared by mixing RTV-SR with nanofillers and then kept for vulcanization at room temperature for 24 h. The RTV-SR composites obtained were characterized with respect to their mechanical, actuation, and magnetic properties. Fourier-transform infrared spectroscopy (FTIR) analysis was performed to investigate the composite raw materials and finished composites, and X-ray photoelectron spectroscopy (XPS) analysis was used to study composite surface elemental compositions. Results showed that mechanical properties were improved by adding fillers, and actuation displacements were dependent on the type of nanofiller used and the applied voltage. Magnetic stress-relaxation also increased with filler amount and stress-relaxation rates decreased when a magnetic field was applied parallel to the deformation axes. Thus, this study showed that the inclusion of iron oxide (Fe3O4) or TiO2 fillers in RTV-SR improves mechanical, actuation, and magnetic properties.


Introduction
Silicone rubber (SR) is frequently used as a composite matrix because it is easily made and has high dielectric and mechanical properties [1,2]. SR is used in a variety of applications such as in actuators [2], strain sensors [3], and coatings [4][5][6]. Among its industrial applications, SR is commonly used to produce actuators and strain sensors [2,3]. With regards to fillers, nanofillers such as carbon nanotubes (CNTs) [7], graphene, few-layer graphene (FLG) [8], nano-titanium dioxide (TiO 2 ) [9], and iron oxide (Fe 3 O 4 ) [10] are often used to produce composites. FLG is used for its high mechanical properties [11,12], TiO 2 for its high dielectric constant [13,14], and Fe 3 O 4 for its ability to orient in a magnetic field and act as a magnetic sensor [15]. From the composite point of view, various types of nanofillers and polymer matrices are mixed to obtain high-performance polymer composites [16] and, in this context, composites based on SR and nanofillers such as CNTs [17], graphene [18], few-layer graphene [19], and carbon black [20,21] are frequently used.
The main role of fillers in rubber composites is to improve mechanical and electrical properties. Several ways have been devised to improve composite properties, for example, (a) using modified (modified or functionalized) fillers [22,23] or polymer matrices to improve interfacial interactions; (b) using hybrid (hybrid or more than one) fillers that RTV silicone rubber (SR) (KE-441, Shin-Etsu, Japan) was used as the polymer matrix for preparing SR composites, and CAT-RM (Shin-Etsu, Japan) was used as the vulcanizing agent. The fillers used to reinforce the SR matrix were Fe 3 O 4 (US Research Nanomaterials, Inc., Houston, TX, USA), FLG (Asbury Mills, Asbury, NJ, USA), and TiO 2 (Alfa Aeser, Ward Hill, MA, USA). The mould release agent was purchased from Nabakem.

Fabrication of Nanocomposites
Composite preparation at room temperature was started by spraying a mould with the mould release agent and leaving it to dry for 3 h. The RTV-SR was placed in a container and mixed manually with a known amount of filler for 10 min. Then, 2 phr vulcanizing agent was added and mixed; the mixture was then poured into a mould and left for 24 h at room temperature to cure. The composites obtained were then removed from the moulds and tested. Samples were designated RTV-SR/FLG, RTV-SR/FLG-Fe 3 O 4 , and RTV-SR/FLG-TiO 2 as described in Table 1. A schematic illustration of the preparation of the RTV-SR composites is shown in Figure 1.   /FLG  100  3  5  10  15  20  2  RTV-SR/FLG-Fe3O4  100  3  5  10  15  20  2  RTV-SR/FLG-TiO2  100  3  5  10  15 20 2 Figure 1. A schematic illustration of the preparation of rubber-based composites.

SEM Micrographs
Composite and filler powder morphologies and surface compositions were studied by SEM (S-4100, Hitachi, Tokyo, Japan) and EDX, respectively.

XRD Analysis
Crystal structures were studied by XRD (D8, Advance Bruker) at room temperature (25 °C) and a scan rate of 10° min −1 .

BET Analysis
Adsorption isotherms were obtained using a BELSORP-max (BEL, Osaka, Japan) at 77 K to estimate BET surface areas of fillers.

Elemental Mapping through SEM
The elemental mapping and morphologies were studied by SEM (s-4100, Hitachi). RTV-SR composites with 20 phr filler were sectioned using a surgical blade to 0.5 mm thick. Samples were also ion-sputter coated with platinum for 2 min. In addition, the presence of Si, O, Fe, C, and Ti elements and their distributions in RTV-SR composites were investigated by X-ray mapping (Horiba EMAX, Tokyo, Japan).

FTIR Analysis
FTIR (Perkin Elmer) was used to identify functional groups of fillers and RTV-SR composites in the transmittance mode over the range 4000-600 cm −1 .

SEM Micrographs
Composite and filler powder morphologies and surface compositions were studied by SEM (S-4100, Hitachi, Tokyo, Japan) and EDX, respectively.

XRD Analysis
Crystal structures were studied by XRD (D8, Advance Bruker) at room temperature (25 • C) and a scan rate of 10 • min −1 .

BET Analysis
Adsorption isotherms were obtained using a BELSORP-max (BEL, Osaka, Japan) at 77 K to estimate BET surface areas of fillers.

Elemental Mapping through SEM
The elemental mapping and morphologies were studied by SEM (s-4100, Hitachi). RTV-SR composites with 20 phr filler were sectioned using a surgical blade to 0.5 mm thick. Samples were also ion-sputter coated with platinum for 2 min. In addition, the presence of Si, O, Fe, C, and Ti elements and their distributions in RTV-SR composites were investigated by X-ray mapping (Horiba EMAX, Tokyo, Japan).

FTIR Analysis
FTIR (Perkin Elmer) was used to identify functional groups of fillers and RTV-SR composites in the transmittance mode over the range 4000-600 cm −1 .  (ESCALAB 250), with a monochromatic Al K σ X-ray source (hν = 1486.6 eV; spot size 200-µm), was used to determine the binding energies of interactions in RTV-SR composites.

Mechanical Properties
Mechanical properties were studied using a universal testing machine (UTM, Lloyd Instruments, West Sussex, UK) at a strain rate of 2 mm/min using cylindrical samples (diameter 20 mm and height 10 mm) at a strain rate of 200 mm/min using dumble-shaped specimens (2 mm thick and with a gauge length of 25 mm).

Hardness
Shore a hardness of the rubber composites were measured using a Westop durometer. A Shore A hardness test is a fast, easy and suitable method for measuring the hardness of a rubber composite. The operation included the application of manual force to the sample and adjusting the testing height to obtain measurements for the specimen with a thickness of around 10 mm.

Actuation Measurements
Actuation displacements were measured using a laser sensor (OptoNCDT 1302) using a 0.1 mm thick electrode with a diameter of 25 mm. The elastomer slab, on the other hand, was 1 mm thick and made of 3 M silicone rubber. The loading of the electrode was 15 phr of nanofiller (FLG, FLG-Fe 3 O 4 , or FLG-TiO 2 ) in RTV-SR.

Magnetic Stress-Relaxation Tests
Stress-relaxation Measurements were obtained in the presence of a magnetic field using a universal testing machine (UTM, Lloyd Instruments, West Sussex, UK) using the same cylindrical samples used for mechanical testing.

Morphology and Purity of Filler Particles
The microstructures of the nanofillers were studied by SEM. FLG had a sheet-like morphology with an ordered distribution of the sheets and a thickness of up to 9-12 nm. A SEM of Fe 3 O 4 revealed the particles were cubic and had a random distribution, whereas TiO 2 particles were spherical with random distribution of the particles and sizes in the range of 17-21 nm. The purities of filler particles are known to influence composite properties. Kumar et al. demonstrated the importance of filler purity in their studies based on SWCNT reinforced rubber composites [44]. We determined the chemical purities of fillers by SEM and EDX (Figure 2a  Fe3O4, FLG, or TiO2. The filler surface properties were determined using adsorption isotherms (Figure 3d-f) and showed volumes of adsorbed gas were related to partial pressure. At a partial pressure of 0.1, BET surface areas for FLG, Fe3O4, and TiO2 were 410 m 2 /g, 5 m 2 /g, and 180 m 2 /g, respectively.

X-ray Diffraction and Adsorption Isotherms of the Fillers
The crystalline states of FLG, Fe 3 O 4 , and TiO 2 were determined by XRD (Figure 3a [46]. Furthermore, XRD showed no evidence of an amorphous phase in Fe 3 O 4 , FLG, or TiO 2 . The filler surface properties were determined using adsorption isotherms (Figure 3d-f) and showed volumes of adsorbed gas were related to partial pressure. At a partial pressure of 0.1, BET surface areas for FLG, Fe 3 O 4 , and TiO 2 were 410 m 2 /g, 5 m 2 /g, and 180 m 2 /g, respectively.

Filler Dispersions by Elemental Mapping in Nanocomposites
The Filler dispersions at a filler loading of 20 phr were determined by elemental mapping (Figure 4). Figure S1 showed that the filler particles were dispersed uniformly in the RTV-SR. Filler particles are indicated by arrows in the high-resolution SEM image.
The homogenous dispersions of nanofillers were confirmed by EDX ( Figure 4). Si, O, and C were densely distributed in all composites (Figure 4a

Compressive Mechanical Properties
The compressive properties of the three composites are shown in Figure 8. Compressive stress and strain were evaluated for FLG (Figure 8a

XPS Analysis
The changes in the surface elemental compositions of RTV-SR composites were assessed by XPS (Figures 6 and 7). Figure 6 shows the XPS scans of RTV-SR, RTV-SR/FLG, RTV-SR/FLG-Fe 3 O 4 , and RTV-SR/FLG-TiO 2 , and their corresponding high-resolution XPS Si2p spectra.    Full survey XPS spectra of RTV-SR and RTV-SR/FLG clearly showed the presence of carbon (C1s), oxygen (O1s), and silicone (Si2s and Si2p), while peaks of iron (Fe2p) [54] and titanium (Ti2s and Ti2p) [55] were observed after incorporating Fe3O4 and TiO2 in RTV-SR/FLG composites (Figures 6 and 7). Furthermore, in the high-resolution XPS spectra of RTV-SR, two main peaks of silicone rubber chains were observed at 102.4 eV (Si-C/O-Si-O ascribed to organic silicone) and 103.5 eV (Si-(O)x, x = 3, 4; inorganic silicone), which represented silicone backbones and side groups, respectively [56,57]. However, no significant changes in the heights or shifts of high-resolution XPS Si2p peaks were observed between composites. The high-resolution XPS spectra of C1s, O1s, Fe2p, and Ti2p are provided in Figure 7.

Compressive Mechanical Properties
The compressive properties of the three composites are shown in Figure 8. Compressive stress and strain were evaluated for FLG (Figure 8a However, compressive moduli fell upon increasing loading from 15 to 20 phr for FLG and the FLG-Fe 3 O 4 hybrid, which we attributed to the aggregation of filler particles at high filler contents and non-uniform filler dispersions [58]. The better properties of the RTV-SR/FLG-TiO 2 hybrid at 20 phr are believed to be due to some synergistic effect between filler particles [24,29].

Tensile Mechanical Properties
Tensile load and extension were investigated for RTV-SR/FLG (Figure 9a), RTV-SR/FLG-Fe 3 O 4 (Figure 9b), and RTV-SR/FLG-TiO 2 (Figure 9c). Tensile loads increased upon increasing filler content. For example, tensile load at a filler content of 5 phr was 6.6 N for RTV-SR/FLG, 6.4 N for RTV-SR/FLG-Fe 3 O 4 , and 5.1 N for RTV-SR/FLG-TiO 2 , while 15 phr increased it to 14.9 N for RTV-SR/FLG, 10.5 N for RTV-SR/FLG-Fe 3 O 4 , and 14.8 N for RTV-SR/FLG-TiO 2 . These increases in tensile load may have been due to the lubricating effect of FLG with increasing filler content in RTV-SR, efficient load-transfer from polymer chains to filler particles, or effective filler networking. Nonetheless, composite hardness increased on increasing filler content ( Figure 10).

Hardness of the Rubber Composites
Reinforcement by nanofillers such as FLG or hybrid filler (RTV-SR/FLG-TiO2 or RTV-SR/FLG-Fe3O4) played an important role in affecting mechanical, electrical, hardness and viscoelastic properties of the rubber composites. Among these properties, hardness was an important property which improved with increasing filler loading in the rubber matrix. Hardness, especially micro-hardness, was important to determine because high microhardness is good enough to break water films on a rubber surface. However, hardness is affected by the high elasticity of the rubber matrix. In order to investigate how soft rubber produced scratches on a glass surface, the hardness of the fillers was considered [59].
Shore A hardness of composites increased with filler content (refer to Figure 10a-c). The variation in hardness between the samples was clear, measured and almost linear as described in Figure 10. The almost linear increases in the hardness of the composites were witnessed with increasing filler content. For example, at a filler loading of 5 phr, Shore A hardness of RTV-SR/FLG, RTV-SR/FLG-Fe3O4, and RTV-SR/FLG-TiO2 were 23, 23, and 24, respectively, and at 20 phr these increased to 30, 32, and 35, respectively, which indicated the fillers had reinforcing and stiffening effects. [9] Furthermore, the hardness of RTV-SR/FLG-TiO2 was greater than those of RTV-SR/FLG and RTV-SR/FLG-Fe3O4 at all filler loadings, which we attributed to the improved polymer-filler interactions and synergistic effects among fillers. The pattern of the results of hardness, especially FLG, was in agreement with ElFaham et al. for graphite-based rubber composites [60]. Rubber composite with a Shore A hardness of 65 or less is rather soft compared to glass [59]. It is possible for

Actuation Measurements of the Rubber Composites
The RTV-SR can be frequently used as an elastomer slab and electrode in actuation due to its higher dielectric properties. In addition to this, RTV-SR has several benefits when used as an elastomer slab, such as easy processing, lower stiffness, easy to vulcanize, and high chemical resistivity. The actuation displacement can be significantly affected by the thickness of the elastomer slab and many other parameters [61].
Actuation displacements of electrodes prepared using rubber composites with 15 phr  Figure 4) at higher filler contents [58], and we suspected the higher tensile strength of RTV-SR/FLG-TiO 2 at 20 phr was due to synergism between FLG and TiO 2 [24,29].

Hardness of the Rubber Composites
Reinforcement by nanofillers such as FLG or hybrid filler (RTV-SR/FLG-TiO 2 or RTV-SR/FLG-Fe 3 O 4 ) played an important role in affecting mechanical, electrical, hardness and viscoelastic properties of the rubber composites. Among these properties, hardness was an important property which improved with increasing filler loading in the rubber matrix. Hardness, especially micro-hardness, was important to determine because high microhardness is good enough to break water films on a rubber surface. However, hardness is affected by the high elasticity of the rubber matrix. In order to investigate how soft rubber produced scratches on a glass surface, the hardness of the fillers was considered [59].
Shore A hardness of composites increased with filler content (refer to Figure 10a-c). The variation in hardness between the samples was clear, measured and almost linear as described in Figure 10. The almost linear increases in the hardness of the composites were witnessed with increasing filler content. For example, at a filler loading of 5 phr, Shore A hardness of RTV-SR/FLG, RTV-SR/FLG-Fe 3 O 4 , and RTV-SR/FLG-TiO 2 were 23, 23, and 24, respectively, and at 20 phr these increased to 30, 32, and 35, respectively, which indicated the fillers had reinforcing and stiffening effects. [9] Furthermore, the hardness of RTV-SR/FLG-TiO 2 was greater than those of RTV-SR/FLG and RTV-SR/FLG-Fe 3 O 4 at all filler loadings, which we attributed to the improved polymer-filler interactions and synergistic effects among fillers. The pattern of the results of hardness, especially FLG, was in agreement with ElFaham et al. for graphite-based rubber composites [60]. Rubber composite with a Shore A hardness of 65 or less is rather soft compared to glass [59]. It is possible for composites with stiff fillers to scratch a glass surface depending upon their respective hardness [59].

Actuation Measurements of the Rubber Composites
The RTV-SR can be frequently used as an elastomer slab and electrode in actuation due to its higher dielectric properties. In addition to this, RTV-SR has several benefits when used as an elastomer slab, such as easy processing, lower stiffness, easy to vulcanize, and high chemical resistivity. The actuation displacement can be significantly affected by the thickness of the elastomer slab and many other parameters [61].
Actuation displacements of electrodes prepared using rubber composites with 15 phr of filler are presented in Figure 11. Actuation displacements and voltages for RTV-SR/FLG, RTV-SR/FLG-Fe 3 O 4 , and RTV-SR/FLG-TiO 2 are shown in Figure 11a-c, respectively. The actuation displacements increased upon increasing input voltage. For example, actuation displacement at 2 kV was 0.07 mm for RTV-SR/FLG, 0.05 mm for RTV-SR/FLG-Fe 3 O 4 , and 0.08 mm for RTV-SR/FLG-TiO 2 . At 12 kV these values increased to 0.66, 0.54, and 0.73 mm, respectively. Notably, actuation displacements were highest for RTV-SR/FLG-TiO 2 at all voltages. In actuation, the stiffness and thickness of the elastomer slab is known to affect the actuation displacement significantly. Moreover, the limiting factors that further affect the actuation performance are electrical conductivity of the electrode, strain amplitude, and nature of the material used in electrode. The actuation performance further depends on the aspects of the flexibility of elastomeric materials that cause slippage at interfacial areas or the formation of cracks at the electrode under stain. The actuation displacement can be further improved if the electrode made of RTV-SR is replaced with conductive polymer such as Nafion in the electrode material [62].
is known to affect the actuation displacement significantly. Moreover, the limiting factors that further affect the actuation performance are electrical conductivity of the electrode, strain amplitude, and nature of the material used in electrode. The actuation performance further depends on the aspects of the flexibility of elastomeric materials that cause slippage at interfacial areas or the formation of cracks at the electrode under stain. The actuation displacement can be further improved if the electrode made of RTV-SR is replaced with conductive polymer such as Nafion in the electrode material [62].

Magnetic Sensing Ability of the Composites
When a viscoelastic rubber composite is deformed it generates internal strains, which relax with time [63]. Magnetic fields influence this stress-relaxation when a magnetic filler is included in a composite [64,65], and magnetic fillers exhibit anisotropic behaviour in

Magnetic Sensing Ability of the Composites
When a viscoelastic rubber composite is deformed it generates internal strains, which relax with time [63]. Magnetic fields influence this stress-relaxation when a magnetic filler is included in a composite [64,65], and magnetic fillers exhibit anisotropic behaviour in the presence of a magnetic field and isotropic behaviour in the absence of a magnetic field. Thus, magnetic filler containing composites exhibited magnetic sensitivity.
In this study, the effect of the magnetic effect on stress-relaxation was studied at different Fe 3 O 4 loadings. Figure 12a shows some typical stress-relaxation curves obtained in the presence and absence of a magnetic field. As shown by the Figure, the stressrelaxation rate decreased when a magnetic field was applied parallel to the deformation axis. Figure 12b shows that the magnetic effect on stress-relaxation values increased with filler loading. Magnetic fields increase filler-filler interactions, and thus, increase magnetic moments of magnetic fillers [66]. It is believed that increasing the Fe 3 O 4 concentration in the SR composite increased the number of magnetic dipoles, and thus, increased the effect of magnetism on stress-relaxation.

R PEER REVIEW
14 of 17 the presence of a magnetic field and isotropic behaviour in the absence of a magnetic field. Thus, magnetic filler containing composites exhibited magnetic sensitivity. In this study, the effect of the magnetic effect on stress-relaxation was studied at different Fe3O4 loadings. Figure 12a shows some typical stress-relaxation curves obtained in the presence and absence of a magnetic field. As shown by the Figure, the stress-relaxation rate decreased when a magnetic field was applied parallel to the deformation axis. Figure  12b shows that the magnetic effect on stress-relaxation values increased with filler loading. Magnetic fields increase filler-filler interactions, and thus, increase magnetic moments of magnetic fillers [66]. It is believed that increasing the Fe3O4 concentration in the SR composite increased the number of magnetic dipoles, and thus, increased the effect of magnetism on stress-relaxation.

Conclusions
In the present work, FLG, Fe3O4, and TiO2 nanofillers were used to prepare composites in an RTV-SR polymer matrix. The effects of these nanofillers on mechanical properties, hardness, actuation and magnetic sensor properties were investigated. The functional groups of nanofillers, RTV-SR, and composites were investigated by FTIR and by measuring binding energies by XPS. Tensile testing showed that load increased with increasing

Conclusions
In the present work, FLG, Fe 3 O 4 , and TiO 2 nanofillers were used to prepare composites in an RTV-SR polymer matrix. The effects of these nanofillers on mechanical properties, hardness, actuation and magnetic sensor properties were investigated. The functional groups of nanofillers, RTV-SR, and composites were investigated by FTIR and by measuring binding energies by XPS. Tensile testing showed that load increased with increasing displacement and tensile strength increased with filler loading up to 15 phr for RTV-SR/FLG and RTV-SR/FLG-Fe 3 O 4 hybrid and up to 20 phr for RTV-SR/FLG-TiO 2 . Actuation displacements increased with applied voltage. At 12 kV, actuation displacements were 0.66 mm for RTV-SR/FLG, 0.54 mm for RTV-SR/FLG-Fe 3 O 4 , and 0.73 mm for RTV-SR/FLG-TiO 2 . Stress-relaxation measurements in magnetic fields showed that stressrelaxation rates decreased when magnetic fields were applied parallel to the deformation axis and increased with Fe 3 O 4 loading.
This study has addressed the use of RTV-SR containing FLG with TiO 2 or Fe 3 O 4 for actuators or magnetic sensors. We studied the effects of fillers on mechanical stiffness, actuation and magnetic sensor properties. The target application of this work is soft materials for actuation and MREs. A composite Shore A hardness of below 65 is considered soft and readily available for soft material applications such as flexible devices, actuators, etc. The Fe 3 O 4 filler used in hybrid with FLG shows a promising magnetic effect at lower filler loadings for use as a magnetic sensor.