Mechanical and electrical properties of alumina

Abstract

The mechanical and electrical properties of natural rubber (NR)–alumina (Al₂O₃) composites crosslinked with dicumyl peroxide (DCP) were investigated. Scanning electron micrographs indicate the interaction of the C–C microstructure and that alumina particles are moderately dispersed in the matrix. The X-ray diffraction patterns indicate that Al₂O₃ is of the β phase polytype that possesses high ionic conductivity. The electrical conductivity of the composites with DCP is greater by nearly three orders of magnitude than that of NR–Al₂O₃ composites without DCP and pure NR. The increase in the electrical conductivity and the mechanical property is caused by the strong C–C bond, the tunnelling phenomenon and the ionic polarisation of the alumina particles acting as the dispersed phase in the elastomer matrix.

Keywords

Electromechanical properties Crosslink agents Tunnelling theory Crosslink density

Introduction

The backbone of natural rubber (NR), i.e. polyisoprene, is related to the polyacetylene backbone through the saturation of every other double bond. Polyisoprene is a potential candidate for materials used in various devices: solar cells, light emitting diodes and field effect transistors.1, 2 Raw elastomers, e.g. NR, have poor properties and need to be reinforced. Reinforcement provides improvement in properties, such as tear strength, abrasion resistance, stiffness and hardness. The solid particles (reinforcing agent or filler) and curing agents, to a large extent, control the mechanical properties of the rubber matrix.3 STR XL is one kind of NR grades which means Standard Thai Rubber type XL or dried NR in an extralight colour slab with a low dirt content of <0·04 wt-% having the chemical structure as shown in Fig. 1 and used as astarting material in this work.

Figure 1

Natural rubber structure

Alumina is a linear dielectric material with ionic polarisation and potential uses as electrolytic capacitors, optical materials and photoluminescents.4 The most common form of crystalline alumina is corundum, which has a rhombohedral Bravias lattice with a space group R-3c (number 167 in the International Tables). Alumina also exists in other phases, namely η, γ, θ and δ theta alumina. All the phases have a structure with a spinel-like Al–O network similar to that found in β-alumina polytypes. The structures of β- and β″-alumina are closely related polytypes; both have the same general stacking pattern of Al–O spinel-like blocks separated by Na–O planes.4 The β- and β″-alumina are high temperature solid electrolytes exhibiting high ionic conductivity.4 These materials have been used in energy storage, alkali–metal thermal to electric conversion cells and gas sensors.5 β″-Alumina, which has higher ionic conductivity, is preferred over that of β-alumina. In general, the properties of Al₂O₃ are low thermal expansion, good thermal shock fracture resistance, low density, high creep resistance, good chemical and thermal stability, high melting point (1828±10°C) and excellent toughness and strength. Previous studies6, 7 reported that alumina can reduce heat, humidity, light, ozone and gamma radiations and crack growth in NR. Al₂O₃ is a potential candidate as the dispersed phase in the NR matrix.8^–10 Sulphur is the most widely used material as curing agent for rubbers. When it is used with an accelerator and an activator at elevated temperatures, thermally stable covalent bonds are formed between the elastomer chains at the carbon–carbon double bonds. Oh and Koenig11 reported the preferred use of peroxide as crosslink agent. An important feature of peroxide as curing agent is the superior strength of the C–C bond compared to the S–S or C–S bond in sulphur curing. The stronger bond strengths result in superior heat and creep resistance at high temperatures compared to sulphur crosslinked systems. Bhowmick et al.12 reported that C–C crosslinking is more resistant than C–S crosslinking and has better compression set properties. The C–C crosslinking is induced by peroxides, which are basically divided into two types: peroxide with a carboxylic acid group (dibenzoyl peroxide) and peroxide without a carboxylic acid group [dicumyl peroxide (DCP)]. Peroxide without carboxylic groups has less sensitivity to acids and higher decomposition temperatures and is suitable for compression. The general advantages of peroxide based crosslinking are high temperature resistance, good elastic behaviour in particular compression sets and no discoloration of the finished product.13 Dicumyl peroxide is an organic compound containing the peroxy group (–O–O–) chain-like structure. Dicumyl peroxide is used as a vulcanising agent for elastomers, a crosslink agent for polyolefins, a polymerising initiator for polyester and a flame retardant for alkenyl aromatic resins (like polystyrene) when combined with organic halogens.

The objective of the present work is to investigate the effect of DCP as crosslink agent on the mechanical and electrical properties of NR and Al₂O₃ composites. The present work also reports on the microstructures and the physical and thermal properties of the NR/Al₂O₃ composites.

Experimental

Materials

The Al₂O₃ (Sigma-Aldrich Chemical Co., Ltd, USA) powder's specific gravity was 3·90 g cm⁻³. The starting NR (STR XL) was supplied by Venus Technology Co., Ltd, Thailand. The STR XL is a dried NR in an extralight colour slab with low dirt content of <0·04 wt-%. The specific gravity, impurity content, Mooney viscosity, initial plasticity and plasticity retention of the NR (STR XL) are 0·92 g cm⁻³, 0·02 wt-%, 62·8 (unitless), 32·0 (unitless) and 81·3 (unitless) respectively. This NR (STR XL) exhibits a high compounded gum tensile strength and good tensile properties and possessed low level non-polymer constituents.1, 7, 14 Dicumyl peroxide (Sigma-Aldrich Chemical Co., Ltd, USA) [C₆H₅C(CH₃)₂]₂O₂, with 99·0 wt-% purity, was used as the crosslink agent.

Instruments

Fourier transform infrared (FTIR) spectra were recorded on a spectrometer (PerkinElmer, model Spectrum One) with spectral resolution of 4 cm⁻¹. The specimens were prepared by mixing with a single crystal potassium bromide.

X-ray diffraction (XRD) patterns were obtained and analysed using an analyser (Bruker, D8 Discover) with a VANTEC-1 detector and a double crystal wide angle goniometer. Scans were taken from 10 to 80° 2θ at a scan speed of 5° 2θ/min in 0·05 or 0·03° 2θ increments using Cu K_α radiation (λ = 0·154 nm). Peak positions were compared with the standard of the Joint Committee on Powder Diffraction Standards files to identify the crystalline phases.

Simultaneous thermal analysis (STA) measurements were taken (Netzsch, model STA 490). The thermal scan was from 27 to 600°C at a heating rate of 10°C min⁻¹ under air atmosphere to determine the vulcanisation behaviour of the composites cured with DCP. The heat of vulcanisation was calculated by integrating the area under the exothermic signal.

Scanning electron microscope (SEM) images were obtained using SEM (model JEOL-5200). Samples of NR, NR/Al₂O₃ and Al₂O₃ powder were mounted on stubs using a carbon paste and sputter coated to ∼0·1 μm gold to improve conductivity. An acceleration voltage of 20 kV with magnifications of ×100 and ×2000 was used.

The electrical properties were obtained and measured using an impedance analyser (HP, model 16451B) with an inductance, capacitance and resistance meter (HP, model 4284 A). The samples were prepared according to ASTM B263-94 for electrical property measurement. Composite samples were prepared as thin discs having a diameter of 25 mm and a thickness of 0·50 mm. The electrical properties were measured at frequencies from 500 to 10⁶ Hz with an ac current of 2 A. The measured electrical resistivity of the composite materials was converted to electrical conductivity as follows15 (1) where ρ_eff is the effective electrical resistivity (Ω m), R is the resistance (Ω), A is the cross-section discs (m²) and L is the thickness (m).

Cumulative and fractional distributions were measured using a particle size analyser (Mastersizer S, model Polydisperse 2·19). The samples were dispersed in a water medium and vibrated in an ultrasonic bath for 20 min.

The curing behaviour was investigated at 160, 170, 180 and 190°C using an oscillating disc rheometer (Techpro-Rheo Tech, ODR121105).

Crosslink density measurement was performed using the toluene swelling method prescribed in ASTM D-6814-02. The vulcanised samples cured at 160, 170, 180 and 190°C were cut into squares (1 cm wide and 3–5 mm thick) and weighed before being swollen in toluene until equilibrium, which normally took 72 h. The crosslink density was determined at room temperature and calculated using the Flory–Rehner equation (2) (3) where ν_e is the effective number of chains in a real network per unit volume (mol cm⁻³), V_r is the volume fraction of polymer in a swollen network in equilibrium with pure solvent, χ₁ is the Flory–Huggins interaction parameter (polymer–solvent interaction parameter for toluene and cis-polyisoprene equal to 0·391 at 25°C) and V₁ is the molecular volume of solvent (toluene), Mw/density or 106·3 (cm³ mol⁻¹). The swelling ratio was calculated from (4) where w₁ is the weight of dry rubber (g), and w₂ is the weight of swollen rubber (g).

Tensile tests were carried out according to the ISO 37 and ASTM D42 procedures, with a type 1 dumbbell specimen in tension mode using a tensile testing machine (Zwick, model Instron 1011) at a constant crosshead speed of 500 mm min⁻¹. Compression moulding and compression set were tested according to ISO 815 on samples of 6 mm thickness and 13 mm diameter. The tests were conducted at the temperature of 23°C and at 55% humidity.

Figure 2

Dumbbell sample for tensile mode (ASTM D42)

Composite sample preparation

Composite samples of NR and Al₂O₃ with DCP were prepared using a two-roll mill (Kodair Seisakusho model R11-3FF). The dispersed phase (Al₂O₃) and the crosslink agent were added in appropriate amounts. The composites were vulcanised for the durations required to obtain 90% cured reaction or t₉₀ as determined from the rheometer curing curves. Three curing conditions were used: adding 1, 3 and 5 wt-%DCP as the crosslink agent at 170°C (the formulations of the composite samples are shown in Table 1).

Table 1

Formulation of composite samples*

Samples	Weight of STR XL/g	Weight of Al₂O₃/g	Weight of DCP/g	Volume fraction of DCP Φ
STR XL_Al_1 wt-%DCP	100	60	1·000	0·0078
STR XL_Al_3 wt-%DCP	100	60	3·000	0·0232
STR XL_Al_5 wt-%DCP	100	60	5·000	0·0380

*Density of STR XL, 0·92 g cm⁻³; density of alumina, 3·90 g cm⁻³; density of DCP, 1·02 g cm⁻³.

Results and discussion

Physical characterisation

The particle size distribution of the DCP shows that the cumulative fraction finer than the size of 90% (d₉₀), 50% (d₅₀) and 10% (d₁₀) is at 1623·33, 963·33 and 311·85 μm respectively. The average particle size of the DCP is 976·16±22·26 μm. Alumina particle sizes corresponding to the cumulative fraction finer than the size of 90% (d₉₀), 50% (d₅₀) and 10% (d₁₀) are at 27·82, 0·16 and 0·06 μm respectively. The average particle size of the alumina is 9·870±0·023 μm. The chemical composition of the alumina powder, as measured by X-ray fluorescence, indicates that it contains 99·79 wt-% of Al₂O₃ and 0·21 wt-% of other oxides.

Figure 3 shows the STA data (differential thermogravimetric–thermogravimetric–differential thermal analysis) of the NR/Al₂O₃ composite crosslinked with DCP and DCP particles respectively. Simultaneous thermal analysis measurements were performed under non-isothermal conditions in order to investigate the vulcanisation behaviour of the uncured composite. Figure 3a shows the percentage of ceramic yield between room temperature and 600°C to be equal to 77·19. The differential thermal analysis thermogram of the composite shows four reaction peaks at temperatures of 170·0, 294·3, 379·1 and 513·6°C under the exothermic signal. The first exothermic peak of the NR/Al₂O₃ composite crosslinked with DCP and vulcanised at 170·0°C occurs at 175·9°C, as shown in Fig. 3b. The general information of DCP halflife occurs at 175·9°C, corresponding to only one peak obtained for the exothermic reaction, as shown in Fig. 3b. The cure time and the cure temperature in a typical application can be related to the halflife of the peroxide. The halflife is the time required for half of the peroxide to decompose at the reaction temperature. Dicumyl peroxide has a short halflife and can initiate curing at a lower temperature. During vulcanisation, the amount of peroxide remaining at any time is the initial concentration of peroxide multiplied by (1/2ⁿ), where n is the number of halflives. The DCP is the peroxide without a carboxylic acid group; it has a compression set at low temperature, is fast curing and decomposes at high temperature. The percentage of DCP remaining is equal to 9·50 as the curing agent decomposition occurs above 175·9°C. The NR/Al₂O₃ composite samples crosslinked with DCP exhibit the viscoelastic behaviour as the temperature is raised above T_g, T_m and T_degrade.15

Figure 3

Simultaneous thermal analysis of samples

The FTIR spectra of the NR, the alumina, the NR/Al₂O₃ crosslinked/uncrosslinked with DCP and the NR with DCP at 27°C are shown in Fig. 4a. The characteristic peaks of NR/Al₂O₃ crosslinked with/without DCP at 170°C are at 3400–3200 cm⁻¹ ν(O–H), 3000–2900 cm⁻¹ ν(C–H) and ( = CH), 2900–2800 cm⁻¹ ν(CH₂), 1800–1600 cm⁻¹ ν(C = C), 1500–1300 cm⁻¹ (CH₃ asymmetric deformation), 1300–1100 cm⁻¹ ν (C–O–C), 1100–740 cm⁻¹ ( = CH bending) and 640–580 cm⁻¹ ν(Al–O), as shown in Fig. 4a. Our FTIR results are consistent with those reported by Ismail et al.16 and Kansal and Laine.17 Figure 4b shows the FTIR spectra of the NR with/without DCP at 170°C: a broaden peak at 3400–3200 cm⁻¹ ν(O–H) of alumina and the NR 3000–2900 cm⁻¹ ν(C–H) and ( = CH), 2900–2800 cm⁻¹ ν(CH₂), 1750 cm⁻¹ and 1031 cm⁻¹ ν(C = C) and ν(C = O), 1152 cm⁻¹ ν(C–O–C), 930 and 740 cm⁻¹ ν( = CH bending) and –(CH₂)_n– (n⩾5) and 640–580 cm⁻¹ ν(Al–O). Figure 4 illustrates the chemical bonding and the functional groups of NR/Al₂O₃ composites through the addition of DCP as the crosslink agent at the curing temperature of 170°C. Dicumyl peroxide as curing agent for NR has one advantage: the resulting network contains only relative stable carbon–carbon bonds as the junction units between chains, providing the crosslinks with high thermal stability, low creep and compression set as well as high reversion and superior heat aging resistance. This is in contrast to the sulphur cured systems, which consist of –C–(S)_x–C– units.

Figure 4

Fourier transform infrared spectra of a NR, alumina (Al₂O₃), NR/Al₂O₃ composite without DCP and NR with DCP, at 27°C and b NR at 170°C, NR with DCP at 170°C, NR/Al₂O₃ with DCP at 170°C and NR/Al₂O₃ without DCP composite at 170°C

The cross-sectional SEM images of the NR (STR XL) and NR/Al₂O₃ composites with 1, 3 and 5 wt-%DCP cured at 170°C are shown in Fig. 5. Figure 5a shows an SEM image of the pure NR. Crosslinking with DCP into the NR/AL₂O₃ composite helps the bridge bonding and, consequently, the mechanical and electrical properties. The higher the percentage by weight of DCP in the NR/Al₂O₃ composites, the stronger and stiffer the bridge bonding in the network, as shown in Fig. 5b–d respectively. The strong and stiff bridge bonding in the network, as indicated by the cross-sectional SEM images of NR/Al₂O₃ composites crosslinked by DCP, is due to the strong C–C bonds (C–C bond strength of 350 kJ).

Figure 5

Images (SEM) of a NR, b NR/Al₂O₃ with 1%DCP cured at 170°C, c NR/Al₂O₃ with 3 wt-%DCP cured at 170°C and d NR/Al₂O₃ with 5 wt-%DCP cured at 170°C

The XRD patterns of the NR, Al₂O₃ and NR/Al₂O₃ composites, with DCP vulcanised at 170°C, are shown in Fig. 6. The XRD patterns of alumina powder and NR–Al₂O₃ composite with DCP cured at 170°C resemble those recorded by the International Center for Diffraction Data (Joint Committee on Powder Diffraction Standards no. 00-042-1458 with the corundum phase form and no. 01-070-0384 with the hexagonal phase form). This observation indicates a partial phase transformation from the corundum phase (rhombohedral) to β- and β″-alumina forms (hexagonal). The structures of β- and β″-alumina are closely related polytypes; both have the same general stacking pattern of Al–O spinel-like blocks separated by Na–O planes.4 The β- and β″-alumina forms exhibit high ionic conductivity; impurity ions (K⁺, Ca²⁺ and Na⁺), as traced by X-ray fluorescence, can migrate through a surplus of cationic sites in the crystal lattice. Furthermore, the XRD pattern of the NR–Al₂O₃ composite crosslinked by DCP suggests that the corundum phase and the rhombohedral form Al₂O₃, which are capable of three polarisations (i.e. electronic, ionic and orientation). This is due to the repositioning of Al³⁺ and O²⁻ ions and the non-centrosymmetric structures: the tetragonal, orthorhombic and rhombohedral forms. In addition, the XRD pattern of NR–Al₂O₃ with DCP shows both crystalline and amorphous phases. The alumina particles in the NR represent the crystalline domains of well packed chains with delocalised charge species dispersed in the amorphous media of low chain ordering. Interparticle resonance tunnelling occurs through the strongly localised states in the amorphous media. In the crystalline region, the NR chains are regularly and densely packed, and there is good interchain overlap. In the amorphous region, the order in the chain arrangement is poor, so the charges are localised in isolated chains.

Figure 6

X-ray diffraction peak patterns of NR, alumina (Al₂O₃) and NR–Al₂O₃ composites with DCP cured at 170°C

Curing characteristics and crosslink density measurement

The curing characteristics of NR/Al₂O₃ with 1, 3 and 5 wt-%DCP and vulcanisation at 170°C are tabulated in Table 2. The results show that the curing time, denoted as t₉₀, is ∼8 min. The effect of DCP as the crosslink agent on the cure characteristics, the optimum curing time and the maximum rheometric torque is considered next. With the incorporation of DCP, the vulcanisation reaction of cured NR–Al₂O₃ composites increases, as indicated by the increase in rheometric torque (more stiffness), curing time and scorch time. This can be explained by the rather good dispersion of the dispersed phase (Al₂O₃) on the NR matrix. It can be assumed that the rheometric torque is related to the crosslink density of the composites.18 The NR–Al₂O₃ composites with 1, 3 and 5 wt-%DCP vulcanisation at 160, 170, 180 and 190°C show swelling ratio and crosslink density characteristics. The NR–Al₂O₃ composites with 1, 3 and 5 wt-%DCP vulcanisation at 170°C show the highest crosslink density. To prove this hypothesis, the crosslink density and swelling ratios of the composites were measured (the data are tabulated in Table 3). At lower and higher vulcanisation temperatures (e.g. 160, 180 and 190°C), the crosslink densities obtained are relatively low. Under- and overcurings lead to reverse reaction and bond breakage (C–C bond breaking).

Table 2

Curing times of STR XL/Al₂O₃ with DCP vulcanisation at 170°C*

Samples	S_max/dN m	S_min/dN m	ΔS/dN m	TS₁/min	TS₂/min	TS₁₀/min	TS₅₀/min	TS₉₀/min
STR XL_Al_1 wt-%DCP	42·53	9·19	33·34	1·04	1·25	1·46	3·17	7·67
STR XL_Al_3 wt-%DCP	72·00	11·05	60·95	0·71	0·83	1·13	2·58	6·04
STR XL_Al_5 wt-%DCP	75·82	11·87	63·95	0·75	0·88	1·13	2·58	7·96

*S_max: maximum torque; S_min: minimum torque; ΔS: difference between maximum and minimum torques; TS₂: scorch time; TS₉₀: optimum cure time; vulcanisation temperature = 170°C.

Table 3

Swelling ratios and crosslink densities of STR XL/Al₂O₃ crosslinked with DCP as measured at room temperature

Samples	Vulcanised temperature/°C	Swelling ratio*	Crosslink densityν_e†/×10⁻⁴ mol cm⁻³	M_c‡/g mol⁻¹	V_r§
STR_XL_1 wt-%DCP	160	1·92	4·96	1854	0·329
	170	3·04	2·17	4249	0·237
	180	3·08	2·11	4354	0·234
	190	4·44	1·08	8530	0·175
STR_XL_3 wt-%DCP	160	2·90	2·35	3911	0·245
	170	2·89	2·36	3900	0·245
	180	2·34	3·48	2643	0·287
	190	4·44	1·08	6975	0·192
STR_XL_5 wt-%DCP	160	2·49	3·10	2968	0·274
	170	2·51	3·06	3008	0·273
	180	3·04	2·16	4263	0·237
	190	2·89	2·38	3869	0·247
STR_XL_Al₂O₃_1 wt-%DCP	160	2·69	2·70	3402	0·260
	170	2·97	3·93	2338	0·301
	180	2·91	2·34	3926	0·245
	190	4·72	0·96	9539	0·167
STR_XL_Al₂O₃_3 wt-%DCP	160	2·68	2·72	3384	0·260
	170	1·96	4·77	1928	0·324
	180	2·19	2·25	4084	0·241
	190	4·08	1·26	7304	0·188
STR_XL_Al₂O₃_5 wt-%DCP	160	2·39	3·32	2767	0·282
	170	1·68	6·29	1462	0·360
	180	3·16	2·01	4583	0·229
	190	3·38	1·78	5177	0·218

*Swelling ratio was calculated using the following equation: swelling ratio = (w₂−w₁)/w₁, where w₂ and w₁ are the weights of swollen rubber and dry rubbers respectively.

†ν_e: crosslink density per unit volume calculated using the Flory–Rehner equation (ASTM D-6814-02).

‡M_c: average molecular weight between the crosslinks.

§V_r: volume fraction of rubber in a swollen network in equilibrium with toluene.

Swelling behaviour

Figure 7 shows the relationship between the volume fraction of NR–Al₂O₃ composites with 1, 3 and 5 wt-%DCP cured at 160, 170, 180 and 190°C in swollen vulcanizate and the crosslink density or swelling ratio determined in toluene according to D 6814-02 following the Flory–Rehner equation and the Flory–Huggins interaction parameter. The swelling ratio of the composites at the equilibrium state in toluene solution decreases as the DCP cured at 170°C percentage increases from 1 to 5 wt-%. The NR–Al₂O₃ composite sample with 5 wt-%DCP cured at 170°C has the highest crosslink density or the lowest swelling ratio. The swelling ratio increases gradually with the vulcanisation temperature and shows an overcuring at high temperature, especially at 190°C. Our results are consistent with the swelling behaviour results obtained by Ismail et al.19 In this work, toluene was used as the only solvent, with the Flory–Huggins interaction parameter equal to 0·391.

Figure 7

Crosslink density of NR–Al₂O₃ with 1, 3 and 5 wt-%DCP composites cured at 160, 170, 180 and 190°C versus volume fraction V_r following Flory–Huggins interaction parameter between toluene and elastomer

Mechanical properties of NR–Al₂O₃ composites crosslinked by DCP at vulcanisation temperature of 170°C

The mechanical properties in tensile mode of the NR–Al₂O₃ composites with 1, 3 and 5 wt-%DCP cured at 170°C are shown in Fig. 8, and the data are tabulated in Table 4. Figure 8 shows the tensile stress–strain curves of the cured composites at 170°C as a function of the percentage of DCP. There is a clear trend of increases in the stiffness and the tensile modulus of elasticity with the increase in the percentage of crosslink agent (DCP). The bond strengths for the rubber crosslinked with DCP (C–C bond of 350 kJ) can be compared to those of sulphur curing (C–S bond of 285 kJ and S–S bond of 155–270 kJ).20, 21 Palys et al. reported that carbon–carbon bonds of DCP crosslinking have higher bond strengths as compared to the lower energy sulphur–sulphur bonds from sulphur vulcanisation.22 The mechanical properties of the NR–Al₂O₃ composites crosslinked by DCP are due to the high bond strengths of C–C bonds; this also explains why DCP generally has better heat aging, higher hardness and modulus, reduced elongation, tear strength and flex fatigue. Moreover, Bhowmick et al.12 showed that C–C crosslinking is mechanically better than C–S crosslinking and has better compression set properties. However, NR–Al₂O₃ composites with a small amount of 1 wt-%DCP cured at 170°C are suitable for various applications, such as engineering devices, actuators and biomimetics, because they still have the highest elongation at break and tensile strength values.

Figure 8

Modulus of elasticity of NR–Al₂O₃ composites with 1, 3 and 5 wt-%DCP cured at 170°C

Table 4

Mechanical properties of STR XL/Al₂O₃ with DCP cured at 170°C as measured at room temperature

Samples	Load at break F/N	Area/mm²	Tensile stress σ/MPa	Tensile strength TS/MPa	Elongation at break El/%	Tensile modulus of elasticity E/MPa	Ductility (%El>5)
STR XL_Al_1 wt-%DCP	95·53±13·87	12·50	8·0345±1·06	7·44±1·06	633·26±114	0·6183	√
STR XL_Al_3 wt-%DCP	29·65±2·17	12·16	2·3864±0·06	2·36±0·08	182·88±113	0·6500	√
STR XL_Al_5 wt-%DCP	27·50±1·74	14·74	1·9045±0·06	1·91±0·08	65·86±10	3·5000	√

The NR–Al₂O₃ composites crosslinked by DCP possess good electrical properties because of the three main mechanisms related to alumina particles as the dispersed phase within the NR matrix:

the stress transfer from the matrix to the non isometric particles

ii.

the partial substitution of soft matrix with stiffer filler

iii.

the segmental immobilisation caused by the interaction of polymer chains with filler surface.23

The stress transfer mechanism depends on the inclusion aspect ratio, its orientation to the applied load and the strength of adhesion, and thus, it is considered as a size independent contribution for the case of sufficiently flexible matrix chains. The second mechanism is independent of particle size as well. On the other hand, the segmental immobilisation mechanism contributes to the overall composite reinforcement with the extent primarily affected by the size of the inclusions, becoming important for submicrometre particles with large surface/volume ratio. In addition, other reinforcement principles are ions migrating out of the alumina particle and the electrochemical reaction.

Electrical properties of alumina particles embedded in NR composites (NR–Al₂O₃) crosslinked by DCP at vulcanisation temperature of 170°C

Figure 9 shows the electrical properties (electrical conductivity and dielectric constant) of NR (STR XL), Al₂O₃ and NR–Al₂O₃ composites with 1, 3 and 5 wt-%DCP cured at 170°C versus the frequency from 500 to 10⁶ Hz, measured at room temperature with an ac current of 2 A. With an increasing percentage of DCP, the electrical conductivity and the dielectric constant increase gradually. The dielectric constant values are 1·542, 15·741, 32·053, 68·385 and 74·383 for the NR, Al₂O₃ and NR–Al₂O₃ composites with 1, 3 and 5 wt-%DCP respectively, as measured at room temperature and at 500 Hz. The electrical conductivity values of the NR, Al₂O₃ and NR/Al₂O₃ composites with 1, 3 and 5 wt-%DCP are 6·517×10⁻⁹, 1·502×10⁻⁷, 6·375×10⁻⁶, 1·101×10⁻⁵ and 1·283×10⁻⁵ Ω⁻¹ m⁻¹ respectively, as measured at room temperature and at 500 Hz.

Figure 9

a electrical conductivity of NR, alumina and NR–Al₂O₃ composites with 1, 3 and 5 wt-%DCP cured at 170°C versus frequency from 500 Hz to 1 MHz measured at 27°C and b dielectric constant of NR, alumina and NR/Al₂O₃ composites with 1, 3 and 5 wt-%DCP cured at 170°C versus frequency from 500 Hz to 1 MHz measured at 27°C

Alumina can improve the electrical conductivity of the composites. In the corundum phase and of the rhombohedral form, Al₂O₃ is capable of three polarisations: electronic, ionic and orientation. This is due to the repositioning of Al³⁺ and O²⁻ ions and the non-centrosymmetric structures (the tetragonal, orthorhombic and rhombohedral forms), including the mechanism of charge species transport occurring by resonance quantum tunnelling like a Drude model response for both high and low frequency regimes. Alumina is a good microparticle conductor used as a filler dispersed randomly into the NR matrix affecting the phenomena involved on both microscopic and macroscopic scales. The conductivity models are based on the continuous insulator–conductor transition called the percolation transition. Our electrical conductivity of the NR–Al₂O₃ composite results corresponds to the theory reported by Sau et al.24 The report explained three main theories for the mechanism of electrical conduction through composites having a random distribution of conductive fillers. The first theory is that the conductive filler forms a few continuous chains (conductive networks) in the rubber matrix. Through this continuous network, charge species (electrons) move from one end to the other under an applied electrical field.25 The second theory is in the electron tunnelling theory; the electrical conduction is believed to take place not only by interparticle contact but also by electrons being able to jump or hop across a gap or tunnel through the energy barrier between conducting elements in the NR matrix. The third theory is in accord with the electric field radiation theory; it is assumed that an emission current is caused to flow by the high electric field being generated between conducting elements separated by a gap of a few nanometres.26

Conclusions

The experimental data show that the NR–Al₂O₃ composite with 1 wt-%DCP as the crosslink agent has the highest tensile strength and elongation at break. The electrical properties, namely dielectric constant and electrical conductivity, of the NR–Al₂O₃ composites crosslinked with 1 wt-%DCP are 32·053 and 6·375×10⁻⁶ Ω⁻¹ m⁻¹ respectively. The electrical conductivity σ of the NR–Al₂O₃ composites with 1 wt-%DCP increases by nearly two and three orders of magnitude as compared to those of NR–Al₂O₃ composites without DCP and pure NR respectively. Our results suggest that a small amount of DCP as the crosslink agent added to the composites can induce strong C–C bond crosslinks. As a result of the C–C bond, the covalent C–C bonds induce high strength, high hardness and high heat aging resistance but poor flex fatigue and tear strength. The bond strength of the rubber crosslinked with DCP (C–C bond of 350 kJ) can be compared to that of the sulphur curing (C–S bond of 285 kJ and S–S bond of 155–270 kJ).20, 21 The carbon–carbon bonds of DCP crosslinking have higher bond strengths as compared to the lower energy sulphur–sulphur bonds from sulphur vulcanisation. Therefore, a small amount of Al₂O₃ particles can be used as filler to absorb loss and to store additional elastic energy within the NR matrix. Furthermore, Al₂O₃ is capable of three polarisations: electronic, ionic and orientation due to the repositioning of Al³⁺ and O²⁻ ions and the non-centrosymmetric structures: tetragonal, orthorhombic and rhombohedral forms. The charge species of Al³⁺ and O²⁻ ions in alumina particles correspondingly assume the role of the tunnelling conductance model in the rubber matrix. Therefore, the obtained NR–Al₂O₃ composites with DCP as the crosslink agent are soft and flexible materials and may be potential candidates for biomimetic actuators and/or artificial muscles.

Footnotes

Acknowledgements

The authors would like to thank the following: the Petroleum and Petrochemical College and the Scientific and Technological Research Equipment Centre, Chulalongkorn University; the Rubber Technology Division, Rubber Research Institute of Thailand, Department of Agriculture; the Faculty of Engineering, Kasetsart University; and the Departments of Materials Engineering and Physics, Kasetsart University, for the use of their analytical equipment. We are also grateful for the grant support from the Small Projects on Rubber (SPR) of the Thailand Research Fund. A. Sirivat would like to acknowledge the financial supports from the Conductive and Electroactive Polymer Research Unit of Chulalongkorn University, the Thailand Research Fund (TRF-BRG), the Royal Thai Government and the Center of Petroleum, Petrochemicals and Advanced Materials.

References

Najidha

, Saxena

, Sreeja

, Unnithan

, Predeep

: Mater. Lett. , 2005, 59, 3431–3436.

Lowery

, Richardson

: ‘Solvent, mechanism and theory in organic chemistry’, 181; 1998, New York, Harper Collins.

Ansarifar

, Azhar

, Ibrahim

, Shiah

, Lawton

JMD

: Int. J. Adhes. Adhes. , 2005, 25, 77–86.

Sutorik

, Neo

, Treadwell

, Laine

: J. Am. Ceram. Soc. , 1998, 81, (6), 1477–1486.

Kummer

, Weber

: US patent no. 3,404,306, 1983.

Vinod

, Varghese

, Kuriakose

: Polym. Degrad. Stab. , 2002, 75, 405–412.

Jakubowickz

: Polym. Degrad. Stab. , 2003, 80, 39–43.

Parthasarathy

, Klingenberg

: Mater. Sci. Eng. R , 1996, R17, 57–103.

Rob

DFM

: Toxicology , 1996, 113, 242–246.

10.

Yamagushi

, Suzuki

: Bull. Chem. Soc. Jpn , 1968, 41, 93–99.

11.

, Koenig

: Polymer , 1999, 40, 4703–4708.

12.

Bhowmick

, Hall

, Benarey

: ‘Rubber products manufacturing technology’, 55; 1994, New York, Marcel Dekker.

13.

Thitithammawong

, Nakason

, Sahakaro

, Noordermeer

JWM

: Eur. Polym. J. , 2007, 43, 4008–4018.

14.

Tardif

, Nishiura

, Hou

: Tetrahedron , 2003, 59, 10525–10539.

15.

Callister

: ‘Materials science and engineering: an introduction’, 7th edn, 806; 2007, New York, John Wiley & Sons, Inc.

16.

Ismail

, Shin

, Yoon

, Shon

, Kim

: Sens. Actuators B , 2008, 129B, 834–840.

17.

Kansal

, Laine

: J. Am. Ceram. Soc. , 1997, 80, (10), 2597–2606.

18.

Pojanavaraphan

, Magaraphan

: Eur. Polym. J. , 2008, 44, 1968–1977.

19.

Ismail

, Nordin

, Noor

: Polym. Test. , 2002, 21, 565–569.

20.

Sanchez

, Myers

: ‘Organic peroxides’, ‘Kirk–Othmer encyclopedia of chemical technology’, 4th edn, 203–310; 1996, New York, John Wiley & Sons, Inc.

21.

‘Technical publication and computer diskette’, ‘Half-life peroxide selection based on half-life’, Atofina Chemicals, Inc., Crosby, TX, USA.

22.

Palys

, Callais

, Novits

, Moskal

: ‘Selection and use of organic peroxides for crosslinking’, Proc. Meet. of the Rubber Division, Chicago, IL, USA, May 1998, ACS, Paper 2.

23.

Kalfus

, Jancar

: Compos. Sci. Technol. , 2008, 68, 3444–3447.

24.

, Chaki

, Khastgir

: J. Mater. Sci. , 1997, 32, 5717–5724.

25.

Pramanik

, Saha