The Effect of Hybridized Carbon Nanotubes, Silica Nanoparticles, and Core-Shell Rubber on Tensile, Fracture Mechanics and Electrical Properties of Epoxy Nanocomposites

The paper investigates the effect of adding a combination of rigid nanoparticles and core-shell rubber nanoparticles on the tensile, fracture mechanics, electrical and thermo-mechanical properties of epoxy resins. SiO2 nanoparticles, multi-walled carbon nanotubes (MWCNT’s), as rigid nanofillers, and core-shell rubber (CSR) nanoparticles, as soft nanofillers were used with bisphenol-A-based epoxy resin. Further, the rigid fillers were added systematically with core-shell rubber nanoparticles to investigate the commingled effect of rigid nanofillers and soft CSR nanoparticles. The resulting matrix will be broadly evaluated by standard methods to quantify tensile, fracture mechanics, electrical, and thermal properties. The results show that the electrical conductivity threshold is obtained at 0.075 wt. % for MWCNT-modified systems. For hybrid systems, the maximum increase of fracture toughness (218%) and fracture energy (900%) was obtained for a system containing 5 wt. % of CSR and 10 wt. % of SiO2. The analysis of the fracture surfaces revealed the information about existing toughening micro-mechanisms in the nanocomposites.


Introduction
Epoxy resins belong to a class of highly cross-linked thermoset polymers used most often with reinforcing fibers in a wide range of composite applications, e.g., automotive, aerospace, and pressure vessels [1]. They show a high specific strength, high modulus, dimensional stability, and low creep. At the same time, the high crosslink density makes them intrinsically brittle materials leading to very low fracture energy. The fracture energy can be increased by adding different modifiers to the epoxy resin, e.g., carboxyl-terminated butadiene acrylonitrile (CTBN) [2,3]. Nevertheless, these modifiers are unfavourable to strength and the glass transition temperature (T g ) [4] of modified systems. A newer type of modifier is core-shell rubber (CSR) nanoparticles. They consist of a rubbery core and an epoxy-compatible shell material, which is only a few nanometers thick. These modifiers-induced the toughening mechanisms of rubber-based modifiers, e.g., cavitation, plastic void growth, and eventually shear yielding [5] together with a better adherence to the matrix due to a modified shell material [6,7]. The addition of rigid fillers like MWCNT's [8], TiO 2 [9], Al 2 O 3 [10], Glass [11], and SiO 2 [12][13][14] can improve the strength and modulus of epoxy nanocomposites while also increasing the fracture toughness without decreasing the glass transition temperature of the nanocomposites. Unmodified epoxy materials have high resistivity and come into the group of insulator polymeric materials. Various fillers can be incorporated in epoxy to reduce their resistivity and can be used for various applications electrical applications. The inherent conductivity of fillers, their aspect ratio, interactions between filler surface and polymer, their dispersion and alignment are critical parameters to obtain the conductivity

Materials and Methods
The EPIKOTE TM Resin L1100 (diglycidyl ether of bisphenol A) liquid epoxy resin and EPIKURE curing agent 943 (amine hardener) produced by Hexion, are used in the present work. The resin and hardener are used in the ratio of 100: 23. The mixed viscosity of the system is 250 mPa.s at 25 • C [22]. The MWCNT's were supplied by Nanocyl, Belgium in the form of 2 wt. % masterbatch in bisphenol-A based epoxy resin. The CSR nanoparticles used in this work are Kane Ace MX267 supplied by Kaneka Belgium NV which has epoxy equivalent weight of 269 g.eq −1 . The material supplied is in the form of a master-batch which was a 37 wt. % concentrate of CSR toughening agent in unmodified liquid epoxy resin based on bisphenol-F based epoxy system, the core of the material is polybutadiene based rubber [23]. The SiO 2 nanoparticles (Nanopox F520) supplied by Evonik Industries in the form of master-batch which contains 40 wt. % of SiO 2 nanoparticles dispersed in bisphenol-F based epoxy system. The epoxy nanocomposites were cured using a two-step curing cycle: (1) 85 • C for 0.5 h, (2) 120 • C for 2 h. Table 1 shows the different systems prepared in this work. Table 1. Nomenclature and composition of bulk epoxy-based composites. In notation (EP_x Y) EP denotes the reference epoxy/hardener system and x represents the wt. % used, and Y represents modifier, i.e. filler and dispersing DGEBA or DGEBF.

Differential Scanning Calorimetry
The DSC experiments were performed on a Q100 (TA instruments, New Castle, DE, USA) system to determine the T g . Firstly, the cured sample material was weighed (~5-10 mg) and placed in a crucible, Nanomaterials 2019, 9, 1057 3 of 11 sealed with lids with the help of crucible sealing press. In the first cycle, the sample was heated from 25 • C up to 200 • C, cooled down to room temperature, and again heated to 200 • C with a heating rate of 10 • C/min.

Dynamic Mechanical Thermal Analysis
In the present study, the storage modulus, the loss modulus, and tan δ of all the bulk samples were measured by dynamic mechanical thermal analysis using a DMA RSA3 machine from TA Instruments, New Castle, DE, USA in 3-points bending mode operating at 1 Hz, on specimens. The glass transition temperature, T g of the bulk epoxy samples was determined by the peak value of tan δ curve. The temperature range was set from 0 • C to 180 • C with a heating rate of 3 • C/min.

Electrical Conductivity
For some applications in anti-static environments, an inherent electrical conductivity of the composite material is required to dissipate the electrical discharge. The volume conductivities of the CNT composites were determined via a 4-point measurement according to DIN EN ISO 3915 (1999). The sample size was approximately 4 × 4 × 8 mm 3 . In all cases, the samples were cut from the middle of dumbbell-shaped specimens and the electrical resistances of the composites were measured. In order to reduce contact resistance, silver conductive paste was applied to the two measuring ends of the sample. The electrical conductivity σ [Siemens/meter] was calculated using the following Equation: where R is the electrical resistance of the specimen; l and A are the sample length and cross-section area, respectively.

Tensile Properties
Tensile tests were conducted at 23 • C on a universal testing machine (Instron Inc., 5500R, Norwood, MA, USA) in a tensile configuration according to standard DIN EN ISO 527-2. Dog-bone shape (ISO 572-2 type 1B) samples were used for the testing. The testing speed was chosen to be 2 mm/min with a 10 kN load cell, a precision sensor-arm extensometer was used to determine the specimen strain.

Fracture Tests
The plane strain fracture toughness (K Ic ) of the composites was determined experimentally at 23 • C by using compact tension (CT) samples under tensile loading conditions according to the Norm ISO 13586 and at strain rate of 0.2 mm/min. The thickness B and the width W of specimens were chosen to be 4 mm and 30 mm, respectively. The samples were tested in a universal testing machine (Instron Inc., 5500R, Norwood, MA, USA). Before, testing a notch was machined and then sharpened by tapping a fresh razor blade into the material, so that a sharp crack was initiated with a length a o (0.45·W ≤ a o ≤ 0.55·W). The fracture toughness K Ic was then calculated by Equation (2), where F is the maximum force observed in the load-displacement curve, and a o is the initial crack length for calculating α = a o /W and f (a o /W ) as follows [24]  The critical stress intensity factor K Ic , the elastic modulus E t and Poisson's ratio ν (~0.35) [25] enable us in calculating the critical energy release rate G Ic

Microscopy Studies
The fractured surfaces of the CT tested specimens were studied with the help of a scanning electron microscope (Zeiss GeminiSEM 300, Carl Zeiss GmbH, Oberkochen, Germany). Prior to scanning, the fractured surface of the samples were sputtered with a thin layer of gold for 120 s using a sputtering device. Transmission electron microscopy (TEM) samples were sectioned using a diamond knife (Diatome, Biel-Bienne, Switzerland) on an ultracut (EM UCT, Leica Microsystems, Vienna, Austria). Ultrathin sections (70 nm) were picked up on copper grids with carbon film. Grids were examined with a Transmission Electron Microscope (Talos F200S G2-Thermofisher-Eindhoven) at 200kV, equipped with a 4K·4K camera One View (Gatan, Paris (France)).

Results
Epoxy cured systems have very low electrical conductivities in the order of 10 −14 to 10 −17 S/cm, however, different fillers have higher conductivities based on their conductive nature. For example, carbon black has conductivity of 10 2 S/cm, graphite 10 5 S/cm and pitch-based carbon fibers have 10 3 S/cm [26]. Nanofillers have exceptionally high electrical conductivity and MWCNTs conductivity lies in the range of 10 4 -10 8 Ω −1 m −1 [27]. For reference epoxy system and the system with 0.05 wt. % of CNT the value to resistance is too high and it's beyond the measuring limit of the equipment. For 0.075 wt. % and 0.1 wt. % the conductivity of composite is measured as 0.00026 S/m and 0.00163 S/m respectively. Table 2 shows the conductivity of all CNT modified systems. The T g values of the epoxy reference system was measured as 120 • C. For all modified systems, the value of T g lies in the range of 120 • C-125 • C, which shows that all the modifiers won't affect the T g of the obtained nanocomposites. Similarly, the T g was also measured using the DMA (tanδ) and for all the systems values lies between 126 • C-130 • C. However, a difference was observed between the T g values in both the techniques which may be due to the difference in the principle of measurements and heating rates used. Figure 1 shows the tanδ variation versus temperature for reference and modified systems.
The tensile properties of the reference and all modified systems are shown Table 3. A modulus of 2700 MPa and a tensile strength of 78 MPa were measured for the EP system. The addition of MWCNT's reduces the tensile strength of the epoxies, which was expected due to possibility of agglomeration during the curing process while elastic modulus remains the same due to the low wt. % of CNT's used. For the EP_20SiO 2 system, the tensile strength and tensile modulus were measured as 90 MPa and 3200 MPa respectively which may be due to the uniform dispersion and effective adhesion between the epoxy matrix and SiO 2 nanoparticles (see Figure 2c). The inclusion of the CSR particles decreases the modulus linearly with increasing filler wt. % and for EP_10CSR the tensile strength and tensile modulus were measured as 55 MPa and 2300 MPa respectively, which may be due to the soft core of the CSR particles. For hybrids, the three different systems were prepared with a concentration of CNT, SiO 2 and CSR fixed at 0.075 wt. %, 10 wt. % and 5 wt. % respectively and excellent dispersion was observed in all the nanocomposites as seen in Figure 2c. Among hybrid systems the highest tensile strength of 82 MPa measured for EP_0.075CNT_10SiO 2 . The tensile properties of the reference and all modified systems are shown Table 3. A modulus of 2700 MPa and a tensile strength of 78 MPa were measured for the EP system. The addition of MWCNT's reduces the tensile strength of the epoxies, which was expected due to possibility of agglomeration during the curing process while elastic modulus remains the same due to the low wt. % of CNT's used. For the EP_20SiO2 system, the tensile strength and tensile modulus were measured as 90 MPa and 3200 MPa respectively which may be due to the uniform dispersion and effective adhesion between the epoxy matrix and SiO2 nanoparticles (see Figure 2c). The inclusion of the CSR particles decreases the modulus linearly with increasing filler wt. % and for EP_10CSR the tensile strength and tensile modulus were measured as 55 MPa and 2300 MPa respectively, which may be due to the soft core of the CSR particles. For hybrids, the three different systems were prepared with a concentration of CNT, SiO2 and CSR fixed at 0.075 wt. %, 10 wt. % and 5 wt. % respectively and excellent dispersion was observed in all the nanocomposites as seen in Figure 2c. Among hybrid systems the highest tensile strength of 82 MPa measured for EP_0.075CNT_10SiO2. The KIc, and GIc of the unmodified epoxy and epoxies modified with different nanofillers were measured in mode I using compact tension (CT) tests. The results are summarized in Table 3. The mean values for KIc and GIc of the unmodified epoxy were determined to be 0.55 MPa·m 1/2 and 0.10  The K Ic , and G Ic of the unmodified epoxy and epoxies modified with different nanofillers were measured in mode I using compact tension (CT) tests. The results are summarized in Table 3. The mean values for K Ic and G Ic of the unmodified epoxy were determined to be 0.55 MPa·m 1/2 and 0.10 kJ/m 2 , respectively. For MWCNT's modified epoxy systems maximum value of K Ic (0.82 MPa·m 1/2 ) and G Ic (0.22 kJ/m 2 ) was achieved at 0.1 wt. % of MWCNT. The maximum values of K Ic = 1.72 MPa·m 1/2 and the G Ic = 1.13 kJ/m 2 were measured for the EP_10CSR system revealing an increase by a factor of~3 and 11.4 above the unmodified epoxy system. For SiO 2 modified systems maximum values of K Ic (1.60 MPa·m 1/2 ) and G Ic (0.70 kJ/m 2 ) was achieved at 20 wt. %. Among hybrid systems maximum value of K Ic = 1.75 MPa·m 1/2 and the G Ic = 1.00 kJ/m 2 were measured for the EP_5CSR_10SiO 2 system. kJ/m 2 , respectively. For MWCNT's modified epoxy systems maximum value of KIc (0.82 MPa·m 1/2 ) and GIc (0.22 kJ/m 2 ) was achieved at 0.1 wt. % of MWCNT. The maximum values of KIc = 1.72 MPa·m 1/2 and the GIc = 1.13 kJ/m 2 were measured for the EP_10CSR system revealing an increase by a factor of ~ 3 and 11.4 above the unmodified epoxy system. For SiO2 modified systems maximum values of KIc (1.60 MPa·m 1/2 ) and GIc (0.70 kJ/m 2 ) was achieved at 20 wt. %. Among hybrid systems maximum value of KIc = 1.75 MPa·m 1/2 and the GIc = 1.00 kJ/m 2 were measured for the EP_5CSR_10SiO2 system. The toughening mechanisms responsible for the increase in fracture toughness due to the incorporation of different nano-fillers can be explained by analyzing the fracture surfaces of the unmodified and modified epoxy composites using SEM technique. The fracture surfaces for the unmodified epoxies, as shown in Figure 3a, for the amine-cured epoxy, show a smooth fracture for inherent brittle epoxy system, which indicates the brittle behaviour of epoxy system during the The toughening mechanisms responsible for the increase in fracture toughness due to the incorporation of different nano-fillers can be explained by analyzing the fracture surfaces of the unmodified and modified epoxy composites using SEM technique. The fracture surfaces for the unmodified epoxies, as shown in Figure 3a, for the amine-cured epoxy, show a smooth fracture for inherent brittle epoxy system, which indicates the brittle behaviour of epoxy system during the fracture process. For CNT modified nanocomposites a uniform dispersion was observed till 0.075 wt. %. At 0.1 wt. % traces of CNT agglomerates were found at the fracture surface as shown in Figure 3b. Debonded CNT's were observed on the fracture surfaces. Therefore, debonding and fiber pull-out were considered to be the major toughening mechanisms linked with the CNT's in epoxy hardener system. The Van der waals forces create an interactive attraction of the CNT's they can drive CNT cluster into agglomerates in the epoxy matrix during curing process even after proper dispersion achieved during the mixing [28]. The presence of agglomeration was responsible for the reduction in the material's toughness and its fracture energy [8]. This may be the sole reason, which was responsible for the CNT's minor contribution on the toughening of brittle epoxy systems. Though the higher concentration of CNT's increases the stiffness of the material but at the same time stress concentration caused by the CNT agglomerates initiated the crack and lead to fracture of the nanocomposites. The fractographic analysis of the fracture surface of SiO 2 modified epoxy resin with the help of SEM and TEM can give an insight into the cause and location of failure, as well as the dispersion state of the particles within the epoxy matrix respectively. Figure 3d shows a close up of the crack surface in a nanocomposite containing 10 wt. % SiO 2 nanoparticles, which reveals that fracture surface appears to be rougher as compared to unmodified system. Further, it shows the presence of voids around few silica nanoparticles based on these observations some nano and micro reinforcing mechanisms can be proposed. Some of these mechanisms include shear yielding, crack deflection, and particle crack pinning as specified by the small tails behind the particles. Similarly, for CSR modified systems river lines and ridges were observed on CSR modified systems along with voids which formed due to the cavitation of core of the CSR nanoparticles see Figure 3c. For hybrid systems, effective dispersion of nanoparticles were observed as seen in Figure 2c,d. The hybrid systems especially EP_5CSR_10SiO 2 system shows rougher fracture surface due to their higher fracture energy as compared to other hybrid systems. Figure 4a reveals the presence of cavitation of CSR particles followed by void growth. It can be observed that silica nanoparticles were uniformly dispersed, and few cavities were observed around silica nanoparticles. The hybridization of silica and CSR interface provide synergy and results in increased fracture toughness. Heish et al [12] proposed that it may occur due to interaction between the particles of different Poisson's ratio, elastic modulus which results in change of stress field in front of crack and provide enhanced degree of plastic deformation, which was previously established by many researchers that stress field interactions may enhance the intensity of plastic deformation mechanisms and extent of plastic zone [29,30]. Figure 4b shows the fracture surface of 5CSR_0.075CNT system showing the features like pull-out of CNT's along with cavitation of CSR particles followed by plastic void growth. Nanomaterials 2019, 9, x FOR PEER REVIEW 7 of 11 fracture process. For CNT modified nanocomposites a uniform dispersion was observed till 0.075 wt. %. At 0.1 wt. % traces of CNT agglomerates were found at the fracture surface as shown in Figure 3b. Debonded CNT's were observed on the fracture surfaces. Therefore, debonding and fiber pull-out were considered to be the major toughening mechanisms linked with the CNT's in epoxy hardener system. The Van der waals forces create an interactive attraction of the CNT's they can drive CNT cluster into agglomerates in the epoxy matrix during curing process even after proper dispersion achieved during the mixing [28]. The presence of agglomeration was responsible for the reduction in the material's toughness and its fracture energy [8]. This may be the sole reason, which was responsible for the CNT's minor contribution on the toughening of brittle epoxy systems. Though the higher concentration of CNT's increases the stiffness of the material but at the same time stress concentration caused by the CNT agglomerates initiated the crack and lead to fracture of the nanocomposites. The fractographic analysis of the fracture surface of SiO2 modified epoxy resin with the help of SEM and TEM can give an insight into the cause and location of failure, as well as the dispersion state of the particles within the epoxy matrix respectively. Figure 3d shows a close up of the crack surface in a nanocomposite containing 10 wt. % SiO2 nanoparticles, which reveals that fracture surface appears to be rougher as compared to unmodified system. Further, it shows the presence of voids around few silica nanoparticles based on these observations some nano and micro reinforcing mechanisms can be proposed. Some of these mechanisms include shear yielding, crack deflection, and particle crack pinning as specified by the small tails behind the particles. Similarly, for CSR modified systems river lines and ridges were observed on CSR modified systems along with voids which formed due to the cavitation of core of the CSR nanoparticles see Figure 3c. For hybrid systems, effective dispersion of nanoparticles were observed as seen in Figure 2c,d. The hybrid systems especially EP_5CSR_10SiO2 system shows rougher fracture surface due to their higher fracture energy as compared to other hybrid systems. Figure 4a reveals the presence of cavitation of CSR particles followed by void growth. It can be observed that silica nanoparticles were uniformly dispersed, and few cavities were observed around silica nanoparticles. The hybridization of silica and CSR interface provide synergy and results in increased fracture toughness. Heish et al [12] proposed that it may occur due to interaction between the particles of different Poisson's ratio, elastic modulus which results in change of stress field in front of crack and provide enhanced degree of plastic deformation, which was previously established by many researchers that stress field interactions may enhance the intensity of plastic deformation mechanisms and extent of plastic zone [29,30]. Figure 4b shows the fracture surface of 5CSR_0.075CNT system showing the features like pull-out of CNT's along with cavitation of CSR particles followed by plastic void growth.    The plain strain dimension of plastic zone size can be quantified by Irwin's model, assuming that the zone was circular and crack occurs in the matrix, by using Equation (5) [31] where KIc is the fracture toughness and σ yt is the tensile true yield stress of the bulk polymer. A plastic zone radius of 2.64 µ m was calculated for EP reference system. The maximum plastic zone size of 52 µ m was calculated for EP_10CSR, and for all other modified systems the plastic zone size falls between these two limits (see Figure 5) The plastic zone was considerably larger than the radius of SiO2 and CSR nanoparticles see Figure 5. Therefore, these particles lie within the plastic zone and favor matrix toughening by events such as cavitation, plastic void growth, crack pinning, pull-out (for CNT's) and crack deflection compelling the material to dissipate more energy before failure. Based on previous work [30,31] Equation (5) gave a good prediction of the size of the plastic deformation zone for the unmodified epoxies. However, for the filler modified epoxy systems, the degree of triaxiality of the stress around the crick tip is notably lower due to the formation of voids. This ultimately reduces the stress necessary for yielding of the epoxy matrix. Hence the size of plastic deformation zone of the particle The plain strain dimension of plastic zone size can be quantified by Irwin's model, assuming that the zone was circular and crack occurs in the matrix, by using Equation (5) [31] where K Ic is the fracture toughness and σ yt is the tensile true yield stress of the bulk polymer. A plastic zone radius of 2.64 µm was calculated for EP reference system. The maximum plastic zone size of 52 µm was calculated for EP_10CSR, and for all other modified systems the plastic zone size falls between these two limits (see Figure 5) Nanomaterials 2019, 9, x FOR PEER REVIEW 9 of 11 modified epoxy systems is significantly larger than the plastic deformation size of unmodified systems [7] [32,33].  The plastic zone was considerably larger than the radius of SiO 2 and CSR nanoparticles see Figure 5. Therefore, these particles lie within the plastic zone and favor matrix toughening by events such as cavitation, plastic void growth, crack pinning, pull-out (for CNT's) and crack deflection compelling the material to dissipate more energy before failure. Based on previous work [30,31] Equation (5) gave a good prediction of the size of the plastic deformation zone for the unmodified epoxies. However, for the filler modified epoxy systems, the degree of triaxiality of the stress around the crick tip is notably lower due to the formation of voids. This ultimately reduces the stress necessary for yielding of the epoxy matrix. Hence the size of plastic deformation zone of the particle modified epoxy systems is significantly larger than the plastic deformation size of unmodified systems [7,32,33].
It can be observed from the normalized graph ( Figure 6) that, for CNT modified systems modulus and fracture toughness increases with increase in CNT concentration but the increase in values is very marginal. For SiO 2 modified systems maximum increase of 3 and 1.2 times was observed for fracture toughness and tensile modulus respectively for EP_20SiO 2 system. Similarly, for CSR modified systems an increase of 3.25 times was noted for EP_10CSR system. Hybrid systems, containing CNT have electrical conductivity~0.0022 S/m with additional increase in fracture toughness values. Normalized map represents different properties for various nanocomposites making it convenient to select a typical system for certain property requirement.  It can be observed from the normalized graph ( Figure 6) that, for CNT modified systems modulus and fracture toughness increases with increase in CNT concentration but the increase in values is very marginal. For SiO2 modified systems maximum increase of 3 and 1.2 times was observed for fracture toughness and tensile modulus respectively for EP_20SiO2 system. Similarly, for CSR modified systems an increase of 3.25 times was noted for EP_10CSR system. Hybrid systems, containing CNT have electrical conductivity ~ 0.0022 S/m with additional increase in fracture toughness values. Normalized map represents different properties for various nanocomposites making it convenient to select a typical system for certain property requirement.

Conclusions
The structure/property relationship of an amine hardener cured bisphenol A-based epoxy modified with CNT's, rigid pre-dispersed SiO 2 nano-particles, CSR nanoparticles, and a hybrid of CSR_CNT, SiO 2 _CNT and CSR_SiO 2 were investigated in terms of tensile, fracture mechanics properties, electrical properties, and thermo-mechanical properties, which further correlated with microstructural features and toughening mechanisms. The dispersion of all the fillers were uniform with few traces of agglomeration in case of CNT modified systems. The elastic modulus and ultimate tensile strength increased linearly with increasing wt. % of SiO 2 nanoparticles due to the relative rigid nature of SiO 2 nanoparticles and excellent adhesion between epoxy matrix and SiO 2 nanoparticles. In the case of EP_CSR nanocomposites, the CSR nanoparticles were well dispersed with no traces of agglomeration. The elastic modulus and ultimate tensile strength were decreased with the addition of CSR nanoparticles, with increase in fracture toughness and fracture energy due to soft nature of rubber core in CSR. For CNT's based epoxy nanocomposites, the electrical conductivity was threshold value was obtained for 0.075 wt. %. Further hybridization with CSR and SiO 2 maintain the value of electrical conductivity with simultaneous improving the tensile and fracture properties. For EP_0.075CNT_10SiO 2 system the increase of 163% was reported for fracture toughness with a minor enhancement of 5% in tensile strength and tensile modulus as compared to reference system. Of all the systems, the hybrids provide overall better performance in terms of tensile, fracture mechanics, electrical conductivity and thermo-mechanical properties as compared to those of single filler systems.