STRAIN-INDUCED CRYSTALLIZATION AND MECHANICAL PROPERTIES OF NBR COMPOSITES WITH CARBON NANOTUBE AND CARBON BLACK
Abstract
The mechanical properties and strain-induced crystallization (SIC) of elastomeric composites were investigated as functions of the extension ratio (λ), multiwalled carbon nanotube (CNT) content, and carbon black (CB) content. The tensile strength and modulus gradually increase with increasing CNT content when compared with the matrix and the filled rubbers with same amount of CB. Both properties of rubber with CB and CNT show the magnitude of each CNT and CB component following the Pythagorean Theorem. The ratio of tensile modulus is much higher than that of tensile strength because of the CNT shape/orientation and an imperfect adhesion between CNT and rubber. The tensile strength and modulus of the composite with a CNT content of 9 phr increases up to 31% and 91%, respectively, compared with the matrix. Differential scanning calorimetry (DSC) analysis reveals that the degree of SIC increases with an increase in CNT content. Mechanical properties have a linear relation with the latent heat of crystallization (LHc), depending on the CNT content. As the extension ratio increases, the glass-transition temperature (Tg) of the composite increases for CB- and CNT-reinforced cases. However, the LHc has a maximum of λ = 1.5 for the CNT-reinforced case, which relates to a CNT shape and an imperfect adhesion with rubber. Based on these results, the reinforcing mechanisms of CNT and CB are discussed.
INTRODUCTION
Since the discovery of carbon nanotubes (CNTs) by Iijima,1 there has been great interest in the area of nanostructured carbon materials, such as nanoelectronics, composite fabrications, and gas storage.2 Extensive work has been done incorporating different types of CNTs, as nanoreinforcements, nanowires, and nanoconductors in polymeric materials to form new composites that possess high mechanical strength and electrical and thermal conductivities.3Although CNTs have been widely used with different kinds of polymers,4–8 very little work has been done on incorporating both CNTs and CBs into rubber.9-11
To achieve the high-stress transfer of composites based on CNTs, a strong bonding is needed between the CNTs and the polymer chains.12,13 However, CNTs can hardly disperse in a matrix because of their nonreactive surface. Thus CNTs ordinarily require treatment for better dispersion before they can be applied. It has been reported that an acid treatment of CNT could improve the process-ability and performance of composites by introducing carboxylic acid groups on the surface of the CNT, which leads to stabilization in polar solvents and helps to covalently link the polymers.14–16 Other possible surface-modification techniques include plasma, thermal, and laser ablations.17,18
In this study, the chemical method used was an acid treatment, and the physical method was an atmospheric–pressure flame plasma (APFP) treatment to modify the CNTs. The APFP treatment is an environmentally friendly technology and can be incorporated in online production. It does not require a vacuum, and its effectiveness has been demonstrated in the treatment of several materials with different shapes and sizes.19,20 The classical methods available for evaluating the degree of cross-linking in elastomers are stress–strain, swelling, and hardness measurements. The differential scanning calorimetry (DSC) is a convenient and rapid technique for the study and an accurate method to investigate the melting and crystallization behavior of elastomer thermal characteristics.921–23 So, we have adopted direct measurements of the thermodynamic transition energy associated with strain-induced crystallization (SIC) for comparison with the strain-induced (SI) crystallinity of CNT- and carbon-black (CB)-reinforced composites. Until now, there have been few published papers on the CNT-reinforced rubber composite using DSC because of the difficulty in conducting measurements. Elastomeric materials are often exposed to different external forces, which induce changes in the dynamic behavior of the material during their practical use. The macroscopic deformation of the polymer network is accompanied by a preferential alignment of the macromolecular chains along the axis of deformation. This chain alignment may lead to the SIC of a deformed network.24–27 Crystallization, when reversibly induced in the deformation process, is of great importance for the elastomeric properties of the network. Thus, the crystallites would presumably act as filler particles, which generally increase the elastic modulus of rubberlike materials significantly. Also, SIC mechanisms of natural rubber filled with nanofillers have been studied using an in situ synchrotron X-ray diffraction technique.28–30
The objectives of this study are to investigate the mechanical properties and determine the relationship between the SIC and the mechanical properties of the elastomeric composites as functions of CNT content, CB content, and the extension ratio (λ). In addition, a quantitative analysis of the composites was achieved using thermogravimetric analysis (TGA).
MATERIALS AND METHODS
The matrix for the experimental work was an NBR (acrylonitrile–butadiene rubber) prepared as a rubber solution by mixing it in a toluene to NBR ratio of 4:1. We also used a multiwalled CNT (CM-100 Ø = 10 ≈ 15 nm; Hanwha Nanotech Co. Ltd. , Seoul, South Korea), and conductive CB (Conductex SC Ultra; Saehan Silichem Co. Ltd., Gyeonggi-do, South Korea) as reinforcing carbon materials and other ingredients of commercial grade quality. The formulation of the test compound is given in Table I.
Generally, entangled CNTs can't be easily dispersed in a polymer matrix. Hence, it is necessary to shorten the length of the CNTs and to get rid of CNT entanglements. To remove impurities, such as metallic catalysts, in the CNTs, 1 g of CNT was chemically treated with 50 mL solution (nitric acid (60%) 1: sulfuric acid (60%) 3) 2 h at 100 °C. It was then filtered and washed with distilled water several times to remove the acid and dried in oven at 100 °C This acid treatment shortens the length of CNTs and introduces carboxyl and hydroxyl groups into their structure.
The plasma treatment of the CNT was carried out with an APFP treatment apparatus (Super Flame 100 Center; A.P.I., Inc., Dalseo-gu, Daegu, Korea). The ratio of compressed air (100 L/min) and C3H8 (4 L/min) in the mixture was 25:1. The appropriate flame distance between the burner port and the particles was 40 mm, and the processing velocity of the burner port was fixed at 50 m/min. The reciprocating treatment was performed twice.
The dispersion of CNTs involves the dissolution of CNTs in a toluene to disentangle the nanotubes that typically tend to cling together and form lumps, making them difficult to process. For this phase, 1 g of CNT was added to 100 mL of toluene using a weighing balance. This solution was sonicated for 1 hour with a mechanical sonicator (Vibra Cell, Sonic & Materials Inc., Newton, CT) capable of vibrating at ultrasonic frequencies (750 W, 20 kHz) to efficiently disperse the CNTs. Finally, to mix the CNT solution and rubber solution with CB, a mixing machine (Ultra-Turrax T-25 digital disperser; IKA Laboratory Equipment, Wilmington, NC) was used for 1 hour at 10 000 rpm. Then, a dry CNT/CB/NBR mixture was obtained by evaporating the solvent off at 80 °C under vacuum.
Tensile sheets were prepared by curing the dried mixtures in the hot press at a temperature of 170 °C, based on the optimum cure time obtained in an oscillating disc rheometer (ODR; Alpha Technologies, Bellingham, WA) measurements. The specimen geometry was a dumbbell shape (60 × 10 × 1 mm) and cut from the cured sheet. The tensile properties were determined using an autograph (Model AG-5000E; Shimadzu Scientific Instruments, Kyoto, Japan) with a testing speed of 50 mm/min.
DSC is an accurate, although limited, method of investigating the melting and crystallization behavior of materials. The details of this experimental technique and its verification can be obtained in ref 27. The degree of SI crystallinity and the Tg of the sample were measured with a DSC Q200 (TA Instruments, New Castle, DE) while maintaining an elongated state. A sample dried completely in a vacuum oven was used for analysis before stretching. The sample was heated from −80 °C under a nitrogen atmosphere to 300 °C at a rate of 10 °C/min. As shown in Figure 1, holders for the stretched samples were machined from brass because of its high thermal diffusivity, high melting point, and low expansion. The degree of the stretch, defined as λ = the length of uniaxially deformed specimen/a relaxed specimen, was chosen to be λ = 1 (relaxed state) and λ = 1.5, 2, 2.5, and 3. As the strip of rubber stretched to a designated length, the holder was rolled up and tight. On average, the size and the weight of the rubber specimen was 1 × 1 × 10 mm before stretching and 12.5 ± 0.2 mg, respectively. Also, during the DSC analysis, three specimens were tested for a single evaluation, and five specimens were used to confirm the result, if needed.



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955
Thermogravimetric analysis was performed using an SDT Q600 (TA Instruments) to evaluate the onset of degradation and the weight loss of the oil, polymer, and residue that remained after the test. The samples were scanned from 30 to 700 °C at a heating rate of 10 °C/min in a nitrogen atmosphere. Typically, four specimens were used for a single evaluation.
RESULTS AND DISCUSSION
Many researchers have been studied the CNT reinforced natural rubber (NR).3,10,11 However, few studies have tested composites enhanced with CNT and CB at the same time. NR has good mechanical and dynamic properties. In spite of that, its electric resistance is comparatively high. In this study, an NBR polymer with a relatively low electric resistance was selected to develop the elastomeric composites for the electric active material, electromagnetic wave-shielding material, and the sensor.31 The results for the tensile strength (σ) and modulus (E) of the composites are presented in Figures 2 and 3 as functions of the CNT and CB content. The tensile strength and modulus gradually increase with the CNT content. The tensile strength ratio (σcomp/σmatrix) and the modulus ratio (Ecomp/Ematrix) of the composites without CB was higher than those of the composites with CB, as shown in Figure 4. There seems to be a higher rate of increase because the matrix without CB (σCB=0 = 2.6 MPa) is weaker than other matrixes with the CB are (σCB=20 = 7.8 MPa; σCB=40 = 15.4 MPa). However, the increasing ratio as a function of CB is similar.



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955
Also, the results indicate that the increasing ratio of the tensile modulus is higher than that of the tensile strength. This is caused by a CNT configuration along the rubber chain at a region of low strain with imperfect adhesion between the CNT and the rubber. Figure 5 shows a schematic of the SIC of the rubber molecules, including the CNT and CB, by its external deformation. Depending on the type of reinforcing fillers, the molecule orientation is different. The case of the CNT clearly shows the molecule orientation because of its aspect ratio (l/d). As the material extends, the separation between the CNTs and the matrix occurs and finally fractures. Because the CNT has a high aspect ratio and can be deformed at an early stage of the extension, the modulus of the composites is more sensitive than the strength of the composites. The tensile strength and the modulus increase up to 31% and 91%, respectively, compared with the matrix at a CNT content of 9 phr.



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955
Figure 6 shows the stress-strain curves of the matrix and several kinds of composites with carbon materials. The tensile modulus indicates the linearly regressed value at the initial region (ɛ = 0.2 ≈ 0.5; □) of the stress-strain curve, and the other modulus represents the value of the slope near the breakage region (○) called the modulus near rupture (MNR) in this study. Our results for the tensile modulus and the MNR are summarized in Figures 7 and 8. As the CNT content increased, the tensile modulus improved greatly, and the CNT reinforcement was larger than that of CB reinforcement with the same content.



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955
However, the MNR of the composites before the breakage step was almost no different from the increase in CNT content compared with the matrix, but the MNR of the composites with CB increases with an increase of content. Therefore, as discussed, CNT facilitates a smooth stress transfer with its large aspect ratio when compared with CB in the region of initial strain, and the tensile modulus is expected to increase significantly by the combination of rubber molecules surrounding CNT.
Because the tensile modulus increases with the filler content, the modulus magnitude of two reinforcing components follows the Pythagorean theorem and can be written as
From the results shown in Figure 7, the k obtained is 1.063 for CNT 6 and CB 20 and 1.017 for CNT 9 and CB 40, respectively. Considering the experimental error, k can be assumed to be 1.
In addition, the tensile strength increases with the filler content, and the strength magnitude of two reinforcing components can be written as
From the results shown in Figure 9, the ϕ obtained is 1.078 for CNT 6 and CB 20 and 1.17 for CNT 9 and CB 40, respectively. Considering the experimental error, ϕ can be also assumed to be 1.



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955
From that, it can be concluded that the reinforcing effects on the tensile modulus and strength by both CB and CNT work simultaneously when both fillers are used together, even though the reinforcing mechanisms are quite different.
To further characterize the interactions among CNT, CB, and the matrix, DSC measurements were conducted. DSC defines the glass transition as a change in the heat capacity as the polymer matrix goes from the glass state to the rubber state. That transition needs the heat, so, in the DSC, the transition appears as a step transition, rather than a peak as might be seen in a crystalline region. Figure 10 shows the typical DSC curve of the matrix. The results for the Tg of the matrix and the composites with an extension ratio of 1.5 are shown in Figure 11. The Tg of the matrix is observed at −34.2 °C and that of CNT-reinforced composites increases with an increasing content. Figure 12 indicates the results for the LHc at a 1.5 extension ratio. The crystallization enthalpy is calculated by integrating the heating power of the crystallization peak in time, divided by the sample mass. DSC analysis reveals that the degree of SI crystallinity of the composites increases with an increase of CNT content. The LHc of matrix is 0.9 J/g. When the CNT content is 9 phr, the LHc is 4.48 J/g, and that of the CB content at 40 phr is 3.7 J/g, whereas that of multiple particles (CNT 9 and CB at 40 phr) is 4.32 J/g. Similarly, the SIC magnitude of two reinforcing components can be written as
where the β obtained is 0.65 ≈ 0.7 for various λ values and filler contents. This represents the different roles of CB and CNT as nucleating agent and physical networking with the matrix that facilitates SIC when those are used at the same time.



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955
As the extension ratio increases, the results of the Tg and LHc are shown in Figure 13 at CNT 6 phr. The Tg of the composite increases with the increase of λ, and the LHc of that shows the maximum at λ = 1.5. In general, the crystalline region is stronger than the amorphous region. The increase of LHc means a reinforcing effect on the composites, and this will improve the mechanical properties. Figure 14 shows that the mechanical properties have a linear relationship with the LHc, depending on the CNT content. The tensile strength and modulus of the composites increase with the increasing CNT content, when compared with the matrix. Also, the LHc shows the same trend and increases up to five times compared with the matrix at a fiber content 9 phr.



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955
The TGA curves and the differential TGA curves (D-TGA) of the composites are shown in Figures 15 and 16, including the curve of the pure matrix. In the TGA curves, the thermodegradation of the composites took place in one step, producing one peak on the D-TGA curves. For all composites, a constant weight was maintained up to 180 °C, and the weight loss occurred between 180 °C and 520 °C, resulting from the thermal decomposition of the rubber, processing oil, and accelerators. The pure matrix was completely decomposed, whereas the composites remained as CNT, CB, and activators as a residue. According to the TGA, the residue of the matrix was 5.7% and those of composites increase with the filler content as shown in Figure 17. With nitrogen gas for purging, the remaining CNT and CB can be identified. Figure 18 shows a scanning electron microscope (SEM) photograph of the composite with CNT 9 and CB 40 after the TGA test. The D-TGA plots show that the degradation temperature of matrix was 455 °C, whereas those of the composite samples are 458 ≈ 464 °C, which indicates that the thermal stability of composites is improved slightly by the incorporation of CNT and CB.



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955



Citation: Rubber Chemistry and Technology 85, 2; 10.5254/rct.12.88955
CONCLUSIONS
From this study, the following conclusions can be drawn:
-
The mechanisms of SIC for CB- and CNT-reinforced rubbers are quite different because of their shapes and that affects their mechanical properties. The effect of CNT on the tensile modulus is much larger than it is on the tensile strength.
-
The tensile strength and modulus increase with an increase of CNT content when compared with the matrix and rubbers filled with same amount of CB. When the amount of CNT is 9 phr, the tensile strength and modulus of the composite increases up to 31% and 91%, respectively.
-
As the extension ratio increases, the Tg of the CNT-reinforced composite increases, and its LHc increases rapidly and reaches a maximum at λ = 1.5. The mechanical properties have a linear relation with the LHc, depending on the CNT content.
-
The reinforcing effects on the tensile modulus and strength by both CB and CNT work simultaneously when both fillers are used together, even though the reinforcing mechanisms are quite different. The LHc as the degree of SIC shows the different roles played by CB and CNT when both are used at the same time.
-
According to the TGA, the residue of the matrix is 5.7% and those of the composites increase with an increase in filler content. The thermal stability of the composites is improved slightly by the incorporation of CNT and CB.

Geometries and materials used for DSC measurements.

Effects of CNT and CB content on the tensile strength.

Effects of CNT and CB content on the tensile modulus.

Effects of CNT and CB content on the tensile strength and modulus ratio.

A schematic of the strain-induced crystallization.

Stress–strain curves of the matrix and composites.

Effects of CNT and CB content on the tensile modulus.

Effects of CNT and CB content on the modulus near rupture (NMR).

Effects of CNT and CB content on the tensile strength.

Typical DSC curve of the matrix.

Effects of CNT and CB content on the Tg, λ = 1.5.

Effects of CNT and CB content on the LHc, λ = 1.5.

Effect of the extension ratio on the Tg and LHc at CNT 6 phr.

Effect of CNT content on the tensile properties and LHc.

TGA curves of the matrix and composites.

D-TGA curves of the matrix and composites.

Effects of CNT and CB content on the residue.

SEM photograph of the composite after TGA.
Contributor Notes