Study on the Use of CTAB-Treated Illite as an Alternative Filler for Natural Rubber.

Fillers are indispensable for rubber composites. Carbon black as an efficient reinforcing filler is most widely used in the rubber industry. However, the utilization of nonrenewable feedstock, energy consumption, and footprint for making carbon black lead to the seeking of alternative substitutes for carbon black, which is of great significance. Here in this work, the possibility of illite, a most common mineral in sedimentary rocks, as an alternative filler for natural rubber (NR) is determined. It is found that pristine illite slows the curing rate and decreases the cross-linking density of NR, which results in the inferior performance of NR. This is associated with the weak filler-rubber interaction, which is a vital factor in deciding the performance of rubber composites. Therefore, illite has been modified using hexadecyl trimethyl ammonium bromide (CTAB), a commonly used cation surfactant, for improving the filler-rubber interaction. The thus obtained C-illite is confirmed to be efficient for (i) enhancing the illite-NR interaction, (ii) improving the dispersion of illite in the NR matrix, and (iii) accelerating the curing process of NR with increased cross-linking density. All of these lead to significantly improved mechanical properties and wear resistance of the C-illite/NR composites, e.g., a 71.88% increase of the modulus at 300% strain compared to the pure NR and a 23.79% reduction of the DIN abrasion volume compared to the NR filled with 40 phr pristine illite. This illustrates the high possibility of CTAB-modified illite with an optimal particle size as a promising alternative filler of carbon black for reinforcing rubbers.

Introduction

Fillers and additives are indispensable and of great concern in the composition of rubber composites with several purposes, such as reinforcement of rubbers, reduction of the material cost, and improvement of the processability of rubbers.[1] Among the abundant kinds of fillers, carbon black (CB) is confirmed to be most efficient in reinforcing rubbers and, therefore, most widely used in the rubber industry.[1,2] However, the energy consumption in the manufacturing process and utilization of nonrenewable feedstock, i.e., the heavy hydrocarbons, for producing carbon black make it unsustainable. Also, the carbon footprint for making carbon black is tremendous due to the partial combustion of heavy hydrocarbons, which release over 2 tons of carbon dioxide for producing 1 ton of carbon black.[3,4] Therefore, alternative feedstock has been considered as substitutes for producing reinforcing filler of rubbers, especially the ones from potential recyclable waste materials and industrial/agriculture byproducts.[5−9] Also, in some formulations of multifunctional elastomers, the total or partial replacement of conventional fillers (e.g., CB) with unconventional ones has been realized.[10−12] On the other hand, studies on the reinforcement mechanism indicate that while the chemical nature of the fillers or additives, which determines the filler–rubber interaction, appears to be vital for reinforcing rubber composites, the size of the fillers or additives is also of prime importance in elastomer reinforcement.[13,14] It is suggested that any kind of fine particles with a desirable small size can reinforce rubbers.[1] In this case, the abundant natural minerals may be another alternative way for providing substitutes of carbon black, which is important for utilization of natural resources and environmental friendliness.[15−17]

It is reported that about 30% of noncarbon black fillers have already been used in the U.S. rubber industry in the 1970s.[18] About 50% of these inorganic fillers are clay minerals. Clay minerals are now recognized as materials of the 21st century since they are abundant, inexpensive, and environment-friendly.[19] Illite is one of the most important clay minerals in sedimentary rocks.[20,21] It makes up approximately 50% of the clay minerals on earth.[22] Consequently, illite is an attractive choice as an alternative filler for rubbers from both environmental and economic aspects.

Illite is a kind of phyllosilicate with an aluminum–oxygen (hydroxyl) octahedral layer sandwiched between two silicon–oxygen tetrahedral sheets. The high aspect ratio of illite nanolayers, especially for completely exfoliated illite, can enhance the mechanical properties and thermal stability of the illite/polymer composite.[23] Hence, using illite as an inorganic filler for polymer composites is very attractive from a commercial aspect. This leads to a number of studies on the application of illite or modified illite in thermoplastic polymer systems with the acquired illite/polymer composites, such as epoxy resin,[23,24] polypropylene,[25,26] polyethylene,[27,28] poly(ethylene oxide),[29] poly(vinylidene fluoride),[30] and so on,[31] showing distinct improvements in their performances. The use of illite as a filler for rubbers is, however, less concerned.

Clay minerals are known to have a lower reinforcing power than carbon black and silica.[18] In this work, to explore the possibility of illite as a substitute filler for the rubber industry, the reinforcement behavior of illite toward natural rubber (NR) is studied. The results show that pristine illite does not reinforce NR remarkably. This may be related to the weak interfacial interaction between illite and NR. To enhance the illite–NR interaction, modification of illite with hexadecyl trimethyl ammonium bromide (CTAB), a commonly used cation surfactant for improving the filler–rubber interaction, was conducted. It was found that the CTAB-modified illite (C-illite) improves the properties of NR evidently compared with the pristine illite.

Results and Discussion

Structural Characterization of CTAB-Modified Illite

Figure a shows the Fourier transform infrared (FTIR) spectra of CTAB, pristine illite, and C-illite. The characteristic peaks corresponding to Al–OH vibrations at 3620, 798, and 779 cm–1; the Si–O–Si bending vibration at 469 cm–1; and the Si–O–Al bending vibrations at 430 cm–1 (where Al is in an octahedral sheet) and 540 cm–1 (where Al is in a tetrahedral sheet) can be found for both illite and C-illite.[32−36] On the other hand, the appearance of peaks at 2925 and 2854 cm–1, attributed to asymmetric and symmetric stretching vibrations of C–H of CTAB, in the spectrum of C-illite indicates the successful incorporation of CTAB into illite. This is further confirmed by X-ray photoelectron spectroscopy (XPS) results, as shown in Figure b, which show increased intensity of the N1s spectrum from 1.43 for illite to 1.52 for C-illite. Meanwhile, the intensity of the K2p spectrum decreased from 2.01 for C-illite to 1.79 for illite. This is attributed to the exchange of potassium ions in the illite interlayer with a quaternary ammonium cation from CTAB,[24] suggesting intercalation of CTAB into illite by cation exchange.

Figure 1

(a) FTIR spectra of CTAB, pristine illite, and C-illite. (b) XPS spectra of illite and C-illite. The insets show the enlarged N1s and K2p spectra. (c) Wide-angle X-ray scattering (WAXS) patterns of pristine illite and C-illite. (d) Thermal gravimetric analysis (TGA) curves of pristine illite and C-illite. The inset shows the weight loss of two main stages.

To confirm intercalation of CTAB into illite, the WAXD patterns of pristine illite and C-illite were recorded and are presented in Figure c. For illite, the (002) and (004) diffraction peaks appear at 8.82 and 17.67°, respectively.[25,37] The (002) and (004) diffraction peaks of C-illite are shifted to 8.60 and 17.24°, respectively, indicating an enlarged interlayer distance of illite after CTAB treatment and confirming the intercalation of CTAB into illite. The amount of CTAB intercalated into illite was estimated by TGA analysis. As presented in Figure d, the weight loss of both illite and C-illite can be divided into two stages. The weight loss of illite below 200 °C is ascribed to the volatilization of water on the surface and in the interlayer of illite, while the weight loss between 200 and 700 °C represents dehydroxylation of illite. It should be pointed out that the weight loss of C-illite in both stages is larger than that of pristine illite. The 0.16% increased weight loss of C-illite below 200 °C may indicate the existence of residual solvent in the illite interlayer. The weight loss of C-illite between 200 and 700 °C increases from 2.39% for pristine illite to 3.83%. This means that totally about 1.44 wt % CTAB has been intercalated into illite, which thermally decomposes in the temperature range between 200 and 700 °C. Even though the decomposition behavior of C-illite is slightly different from that of pristine illite, their effect on the thermal stability of the filled NR composites is almost the same. As presented in Figure S1 and Table S1 of the Supporting Information, both the illite-30/NR and C-illite-30/NR vulcanizates exhibit a slightly improved thermal stability.

Interfacial Interaction of NR with Illite or C-Illite

The purpose of modification of illite with CTAB is to enhance the interaction between illite and NR, and CTAB is used since it is frequently used for improving the filler–rubber interaction. Therefore, the filler–rubber interaction of illite/NR and C-illite/NR vulcanizates is first characterized. This is actually a hard task due to the problem of isolating the filler–rubber interaction from other possible influencing factors, such as the rubber cross-link and filler networks.[38−42] On studying the interactions between different carbon blacks and varied rubbers, Ayala et al.[39,40] found that the slope of the stress–strain curve under static measurements in a relatively linear region, namely, σ = (σ2 – σ1)/(λ2 – λ1) with λ usually ranging from 1 to 3, can better reveal the carbon black–polymer interaction. On the other hand, the ratio of storage modulus measured under double-strain amplitudes at 1% (G1′) and 25% (G25′), i.e., η = G1′/G25′, is a good indicator of filler–filler interaction.[43] Therefore, the interaction parameter I = σ/η is expected to eliminate the effect of filler–filler interaction and, thereby, accentuate the filler–rubber interaction. Figure a shows the plots of σ, obtained from the slope of the stress–strain curve in a relatively linear region with λ between 1.5 and 3, against the filler content (presented in Figure S2 of the Supporting Information). One can clearly see that σ of illite/NR vulcanizates increases steadily with increasing filler content. σ of illite/NR with 50 per hundred parts of rubber (phr) illite is about 1.88 times higher than that of illite/NR with only 10 phr illite, indicating increased illite–NR interaction with increasing illite content. Furthermore, the η value of illite/NR vulcanizates, see Figure b, increases very slightly from 1.11 to 1.17 on increasing the illite content from 10 to 50 phr, which is only about 1/2 the value of a carbon black/NR system with 45 phr carbon black.[40] This implies a less pronounced filler–filler interaction of illite/NR compared to the carbon black/NR system, indicating better dispersion of illite in the NR matrix. This is related to the different surface properties of illite and carbon black. The abundant hydroxyls existing on the carbon black surface result in their serious aggregation in nonpolar rubbers, which amplifies the effect of the filler–filler interaction.[41,42,44] For illite, there are only a few silicon hydroxyls and aluminum hydroxyls in its edge planes, leading to their weak aggregation in nonpolar NR even at high loadings.[45,46] This is an advantage of illite when used as a filler for polymers. The steadily increasing σ and the almost constant η also result in an increase of the interaction parameter I of illite/NR, as presented in Figure c, which illustrates an improved filler–rubber interaction.

Figure 2

(a) σ obtained from the slope of the stress–strain curve in a relatively linear region with λ between 1.5 and 3, (b) filler network factor (η), and (c) interaction parameter (I) of illite/NR and C-illite/NR vulcanizates with different filler additions.

For the C-illite/NR vulcanizates, it can be seen from Figure a that a ca. 200% increase of σ has further been achieved with respect to illite/NR with the same filler loading. At the same time, the η decreases slightly after the CTAB treatment of illite (Figure b), reflecting a further improved dispersion of C-illite in the NR matrix compared to illite. This demonstrates that modification of illite with CTAB can indeed decrease the filler–filler interaction while increasing the interaction between C-illite and NR effectively (Figure c). This can be understood in the following way. Incorporation of the long alkyl chain of CTAB makes C-illite more organophilic, which improves the compatibility between rubber and C-illite. Moreover, incorporation of CTAB into illite increases the interlayer spacing of illite as confirmed by the WAXD results. The larger interlayer spacing favors the insertion of rubber chain segments into illite, which improves the interaction between C-illite and NR. The increased filler–rubber interaction enhances the dispersity of the filler in the rubber matrix as confirmed by the scanning electron microscopy (SEM) observation (Figure S3), which suppresses the formation of a filler network and consequently reduces the filler–filler interaction.

This reduced filler–filler interaction is further demonstrated by the weak Payne effect of C-illite/NR with respect to the illite/NR compounds, as presented in Figure . As is well-known, the filler network plays an important role in the stress of filled rubbers mainly at low strains.[38] From Figure , we can see that the storage modulus (G′) at low strains increases with filler content for both illite/NR and C-illite/NR systems, indicating the increase of aggregation, i.e., formation of a filler network with increasing filler content. The G′ of C-illite/NR at low strains is, however, always smaller than that of illite/NR with the same loading. This unambiguously demonstrates the effect of CTAB on the suppression of illite aggregation. Actually, the discrepancy of G′ at low (1%) and high (100%) strains, namely, ΔG′ = G′ (at 1%) – G′ (at 100%), is frequently used to reflect the filler network. The larger the ΔG′, the stronger the filler network, which has been proved in other elastomer systems.[47] The ΔG′ for illite/NR with 40 and 50 phr illite was calculated, as shown in Figure , to be 51.29 and 57.59 kPa, respectively, reflecting increased filler network strength. The ΔG′ for C-illite/NR with 40 and 50 phr C-illite was calculated to be 44.65 and 47.13 kPa, respectively, which are clearly smaller than those of the illite/NR systems. The decreased ΔG′ after CTAB modification of illite supports the conclusion that CTAB improves dispersion of C-illite in the NR matrix.

Figure 3

Strain-dependent storage modulus (G′) of uncured illite/NR and C-illite/NR compounds with filler loadings of 40 and 50 phr.

Taking into consideration the fact that the contribution of the filler network to stress can be ignored at higher strains because of its poor deformability, another method has been developed to estimate the filler–rubber interaction with the exclusion of the cross-linking density of the rubber matrix.[48] The filler–rubber interaction is expressed by the density of chain segments introduced by the interfacial phase, i.e., Fint, as presented in the following equationwhere σ is the stress of the vulcanizates, R is the universal gas constant, T is the absolute temperature, λ is the tensile ratio, and Ve is the cross-linking density. According to the σ values obtained from stress–strain curves shown in Figure S2 and the Ve values obtained from the vulcanization characteristics of illite/NR and C-illite/NR composites (cf. Figure and Table ), the plots of Fint as a function of strain are presented in Figure . It is clear that Fint exhibits a very high value at very low strains and decreases quickly as the strain increases to about 50%. This should be ascribed to the effect of the filler network. Fint increases gradually with increasing strain, which means that the interaction of NR with illite or C-illite becomes stronger. This is because the filler–rubber interaction relies on the deformation of the rubber network. It should be pointed out that Fint also increases with increasing filler content. This is associated with the fact that augmentation of filler loading facilitates the introduction of more rubber chains or chain segments into the interfacial phase, thus causing an improvement of the interaction between NR and illite or C-illite. In addition, the values of Fint are obviously higher for C-illite/NR than for illite/NR at the same amount of filler and strain. This demonstrates the improvement effect of CTAB on the filler–rubber interaction.

Figure 5

(a) Vulcanization curves for illite/NR and C-illite/NR compounds; the inset shows the variation of the optimum curing time (t90) with filler dosage. (b) Cross-linking density of illite/NR and C-illite/NR vulcanizates.

Table 1

Tear Strength, DIN Abrasion Volume, Shore A Hardness, and Cross-Linking Density of Different Composites

sample	tear strength (kN/m)	DIN abrasion volume (mm3)	Shore A hardness	cross-linking density (10–4 mol/cm3)
unfilled NR	27.4 ± 0.8	184 ± 16	37 ± 0.5	1.018 ± 0.015
illite-10/NR	27.4 ± 0.5	285 ± 12	36 ± 0	0.915 ± 0.008
illite-20/NR	28.5 ± 0.3	290 ± 11	36 ± 0.4	0.934 ± 0.001
illite-30/NR	29.7 ± 0.9	287 ± 3	39 ± 0.5	0.935 ± 0.001
illite-40/NR	30.6 ± 0.6	311 ± 9	40 ± 0.4	0.971 ± 0.011
illite-50/NR	29.7 ± 0.7	327 ± 17	41 ± 0	0.994 ± 0.008
C-illite-10/NR	33.5 ± 0.9	220 ± 8	43 ± 0	1.290 ± 0.011
C-illite-20/NR	34.3 ± 0.8	227 ± 13	44 ± 0	1.318 ± 0.008
C-illite-30/NR	36.0 ± 0.7	238 ± 4	45 ± 0	1.362 ± 0.011
C-illite-40/NR	37.4 ± 0.8	237 ± 9	48 ± 0	1.446 ± 0.011
C-illite-50/NR	39.8 ± 0.4	238 ± 8	49 ± 0.4	1.475 ± 0.010

Figure 4

Strain-dependent Fint of NR vulcanizates containing different amounts of illite or C-illite.

Vulcanization Characteristics of Illite/NR and C-Illite/NR Composites

It is well-known that filler–rubber interaction is one of the determinants of torque changes for rubber composites during vulcanization.[49,50] As confirmed in the previous section, modification of illite by CTAB enhances the interaction between illite and NR, and the vulcanization characteristics of illite/NR and C-illite/NR composites were then compared. Figure a shows the curing curves of the NR composites with either illite or C-illite and the optimum curing time (t90) as a function of filler dosage (inset). From the curing curves shown in Figure a, one can see that there is only a negligible change for the minimum torque (ML) of the NR composites with different amounts of illite and C-illite. This is reasonable since it is closely related to the rubber matrix. However, the torque increases with curing time, indicating the occurrence of cross-linking. Different evolutions of the torque for the illite/NR and C-illite/NR composites with time reveal the varied curing behavior. For the illite/NR composites, it is clear that the torque increases slower than that of the neat rubber in the first 500 s, implying a slower curing rate of illite/NR systems compared with NR. The torque of NR reaches, however, a plateau after being cured for 600 s. This implies that cross-linking is completed. On the other hand, the torque for illite/NR composites continues to increase. The maximum torque (MH) of illite/NR composites with the illite content higher than 30 phr even exceeds that of NR; see Figure b. One may attribute the higher torque to a higher degree of curing. This is actually not the case since the cross-linking density of all illite-filled NR composites is calculated to be lower than that of the NR (see Figure b and Table ). These results lead to the conclusion that illite slows the curing rate and decreases the curing degree of NR. The slowed curing rate of illite/NR systems is also illustrated by the increased t90 of illite/NR, which also increases steadily with the increase of the illite content. In this case, the enhanced torque of illite/NR composites with more than 30 phr illite is the synergistic effect of filler addition and rubber network formation.

The influence of C-illite on the curing behavior of NR is quite different from that of illite. First, the cure of C-illite/NR composites is shorter than that of their illite/NR counterparts. The corresponding t90 of C-illite/NR shows an almost negligible variation with increasing C-illite addition. The t90 of the C-illite-50/NR compounds is only about half the value of that of the illite-50/NR compounds. This is attributed to the ammonium groups of CTAB, which are known to be beneficial for the curing reaction of rubber compounds and increase the vulcanization rate.[51−53] Taking this into account, C-illite can act as a vulcanization accelerator for improving vulcanization efficiency and reducing energy consumption. Second, the MH of C-illite/NR compounds has a significant increase compared to that of illite/NR. As the amount of the filler increases, the value of MH of C-illite/NR increases continuously as well, resulting in a widened difference between the MH of C-illite/NR and illite/NR (Figure b). As reported by Lei et al.,[38]MH depends mainly on the cross-linking density of the composite and the filler–rubber interaction. While enhanced filler–rubber interaction has been confirmed in the previous section, the increased cross-linking density of C-illite/NR has also been revealed in Figure b and Table . It is clear that the cross-linking density of C-illite/NR increases monotonously with the increasing loading of C-illite, which is attributed to the introduction of abundant physical cross-link points by C-illite sheets.

Table 2

Tensile Strength, Modulus at Different Strains, and Elongation at Break of Different Composites

sample	tensile strength (MPa)	modulus at 100% (MPa)	modulus at 300% (MPa)	modulus at 500% (MPa)	elongation at break (%)
unfilled NR	14.7 ± 0.6	1.3 ± 0.3	2.2 ± 0.3	6.0 ± 0.8	610 ± 39
illite-10/NR	12.2 ± 0.3	0.7 ± 0.2	1.6 ± 0.3	4.5 ± 0.8	645 ± 20
illite-20/NR	14.8 ± 0.5	0.7 ± 0.2	1.9 ± 0.2	6.3 ± 0.4	636 ± 32
illite-30/NR	16.3 ± 0.6	0.8 ± 0.3	2.4 ± 0.3	8.0 ± 0.4	625 ± 21
illite-40/NR	16.7 ± 0.6	1.3 ± 0.3	3.2 ± 0.2	9.2 ± 0.1	624 ± 28
illite-50/NR	16.2 ± 0.6	1.4 ± 0.3	3.9 ± 0.2	11.2 ± 1.0	579 ± 23
C-illite-10/NR	19.9 ± 0.3	1.0 ± 0.3	2.5 ± 0.2	9.5 ± 0.4	607 ± 19
C-illite-20/NR	20.8 ± 0.5	1.2 ± 0.3	3.3 ± 0.3	11.9 ± 0.6	592 ± 6
C-illite-30/NR	21.1 ± 0.5	1.2 ± 0.3	3.9 ± 0.3	13.2 ± 0.6	596 ± 26
C-illite-40/NR	22.0 ± 0.7	2.0 ± 0.3	5.5 ± 0.3	17.0 ± 0.8	555 ± 17
C-illite-50/NR	21.6 ± 0.4	2.3 ± 0.2	6.1 ± 0.2	17.6 ± 0.2	554 ± 5

Mechanical Properties of Illite/NR and C-Illite/NR Composites

It should be pointed out that both the filler–rubber interaction and the cross-linking density of the rubber have a significant influence on the ultimate mechanical properties of rubber composites. The strengthened interfacial interaction facilitates the stress transfer from the rubber matrix to the filler, which helps in improving the mechanical performance of the composites. Therefore, variations of mechanical properties as a function of filler loading for both illite/NR and C-illite/NR vulcanizates were studied and are presented in Figure . The related data can be found in Tables and 2. From Figure a and Tables and 2, one can find that a low illite loading, e.g., less than 30 phr, exhibits no evident reinforcement effect on the tensile and tear strength of the NR. This is different from C-illite, which shows an evident effect even with only 10 phr C-illite. C-illite improves the mechanical properties of composites more effectively than illite due to the increased interfacial interaction and the cross-linking density of the NR. The incorporation of 40 phr pristine illite into NR leads to an increase of the tensile strength from 14.7 MPa for NR to 16.7 MPa for illite-40/NR and the tear strength from 27.4 kN/m for NR to 30.6 kN/m for illite-40/NR. A further increase of the tensile and tear strengths to 22.0 MPa and 37.4 kN/m, respectively, was achieved for C-illite-40/NR. It is noted that the difference in tensile strength between the illite/NR and C-illite/NR systems remains almost unchanged when the filler loading is higher than 20 phr while that of tear strength tends to increase with a further increase in the filler content.

Figure 6

(a) Tensile strength and tear strength, (b) modulus at 300%, (c) Shore A hardness, and (d) DIN abrasion volume of illite/NR and C-illite/NR vulcanizates.

From Figure b and Table , it can be seen that a low loading of illite into NR leads to a decrease of the modulus at different strains. For example, the modulus at 300% decreased from 2.2 MPa for NR to 1.6 MPa for illite-10/NR. But the modulus increases steadily with increasing illite amount. A maximum value of 3.9 MPa at 300% was reached for illite-50/NR. Similar to the case of tensile and tear strengths, the modulus at 300% increases continuously with C-illite. A value of 5.5 MPa was already achieved for C-illite-40/NR, which further increased to 6.1 MPa when the loading of C-illite was 50 phr. These values are somewhat lower than those of NR filled with 30 phr carbon black with the particle size ranging from 26 to 100 nm as presented in Table S2 of the Supporting Information. They are however much higher than those of NR filled with 30 phr Z200MP silica with the particle size ranging from 15 to 20 nm (see Table S2). The tensile strength of NR filled with 30 phr C-illite is even higher than that of the NR filled with 30 phr Si69-modified Z200MP silica (21.1 vs 15.4 MPa). Moreover, we recall that the filler size plays an important role in elastomer reinforcement.[13,14] This has actually been verified by further improvement in the properties of an N330 carbon black (particle size of 26–30 nm)-filled system compared to an N774 carbon black (particle size of 61–100 nm)-filled one, as presented in Table S2. Taking all of these into account, C-illite with the desired size is expected to be an efficient filler for reinforcing NR.

It is worth noting that the augmentation of the cross-linking density is also accountable for the increase of the modulus because it can also influence the deformability of the rubber network. This is usually illustrated by the Shore A hardness, which is related both to the cross-linking density of rubber composites and to the filler–rubber interaction. From the plots of the Shore A hardness and the cross-linking density against filler dosage shown in Figures c and 5b, as well as the data summarized in Table , we can see that a lower cross-linking density leads to a low Shore A hardness of illite/NR compared to C-illite/NR. It should be pointed out, however, that the Shore A hardness of illite/NR becomes higher than that of unfilled NR when the dosage of illite is over 30 phr even though the cross-linking density of illite/NR is always smaller than that of the neat NR (see Table ). This reflects a vital role of the interaction between NR and illite, which increases with increasing illite content, on the Shore A hardness of the composites. For the C-illite/NR system, the variation trend of the Shore A hardness with the filler dosage is basically the same as that with the cross-linking density shown in Figure b. It is more clear from Figure that the exponential index is 1 for the fitted curve (the Shore A hardness vs the cross-linking density of C-illite/NR composites), which is consistent with that of a previous study.[54] This is because the Shore A hardness is equivalent to the modulus by definition and the modulus is proportional to the cross-linking density according to the theory of affine deformation.

Figure 7

Linear relationship between the Shore A hardness and the cross-linking density for C-illite/NR composites.

As wear resistance is an all-important index to determine the working life of rubber products, the DIN abrasion volume of illite/NR and C-illite/NR was compared; see Figure d and Table . It is widely accepted that the variation of abrasion for the same rubber matrix depends mainly on the resistance of rubber to fracture, which is decided by the dispersion of the filler and the modulus of the composites.[55] Enhanced filler dispersion as illustrated by Figure b and the increased modulus displayed in Figure b result in superior wear resistance of C-illite/NR with a smaller DIN abrasion volume compared to illite/NR. Moreover, the wear resistance of illite/NR shows a decreasing trend on increasing the filler loading to over 30 phr. The deterioration of the wear resistance for illite/NR with increasing illite content may be attributed to illite aggregation. On the contrary, the wear resistance of C-illite/NR remains almost unchanged in the loading region of C-illite from 30 to 50 phr, again indicating better dispersion of C-illite in the NR matrix.

Conclusions

The effect of illite before and after CTAB treatment on the filler–rubber interaction, vulcanization characteristics, and mechanical properties of NR have been studied in detail, and the following conclusion can be drawn from the obtained results.

Insertion of CTAB molecules into interlaminated illite by cation exchange enhances the interaction between illite and NR and improves the dispersion of C-illite in the NR matrix. As a result, the reinforcement of C-illite toward NR is superior to that of pristine illite. For example, the increase of tensile and tear strengths from 16.7 MPa and 30.6 kN/m for illite-40/NR to 22.0 MPa and 37.4 kN/m for C-illite-40/NR, respectively, has been achieved. Especially, a modulus at 300% of 6.1 MPa has been obtained for C-illite-50/NR, which is about three times the value for pure NR.

The study of the vulcanization characteristics of illite/NR and C-illite/NR composites shows that the addition of illite into NR not only slows down the curing rate but also reduces the cross-linking density of illite/NR. On the contrary, the curing time of C-illite/NR is much shorter than that for both the pure NR and the illite/NR composites, and the corresponding t90 of C-illite/NR shows an almost negligible variation with increasing C-illite addition, indicating that C-illite can act as a vulcanization accelerator for improving the vulcanization efficiency and reducing energy consumption. Taking all of these into account, C-illite with the desired size is highly expected to be an efficient filler for reinforcing NR.

Experimental Section

Materials

Natural rubber (NR, SCR-WF) was purchased from Hainan Rubber Industry Group Co., Ltd., China. Illite was kindly supplied by Chengde Renhe Mining Co., Ltd., Hebei, China. The specific surface area of illite was 21 m2/g (determined by gas adsorption using the Brunauer–Emmett–Teller (BET) method), and the grain size was ca. 10 μm in diameter. The SEM micrograph and energy-dispersive spectra (EDS) of pristine illite are shown in Figure S4. Hexadecyl trimethyl ammonium bromide (CTAB) was obtained from Sinopharm Chemical Reagent Co., Ltd. Other chemicals were commercial analytical reagents and used as received. All of the rubber ingredients were of industrial grade and used as received.

Modification of Illite by CTAB

Illite was dispersed in ethanol at a concentration of 50 mg/mL, into which CTAB was incorporated with an illite/CTAB mass ratio of 20:1 according to ref (25). Then, the mixture was magnetically stirred for 5 h at 60 °C, followed by repetitive centrifugation and washing with ethanol to remove the dissociative CTAB. Finally, the products, denoted as C-illite, were dried in vacuum at 60 °C for 24 h. For direct comparison of the reinforcement behavior of C-illite and illite toward NR, pure illite was treated with exactly the same procedures as C-illite, including stirring at 60 °C for 5 h, centrifugation, washing with ethanol, and drying in vacuum at 60 °C for 24 h.

Preparation of Illite/NR and C-Illite/NR Composites

The basic formulation of rubber composites is as follows: zinc oxide (ZnO), 5 per hundred parts of rubber (phr); 2-mercaptobenzothiazole (M), 1 phr; stearic acid (SA), 2 phr; N-isopropyl-N′-phenyl-4-phenylenediamine (4010NA), 2 phr; and sulfur (S), 3 phr. The illite/NR and C-illite/NR composites were fabricated by incorporating different amounts of illite and C-illite into the rubber composites, respectively. Typically, an appropriate amount of NR with activators ZnO and SA, accelerator M, and antioxidant 4010NA was added to a two-roll mill to first obtain the evenly dispersed rubber composites, then the desired amount of illite or C-illite was added, followed by addition of vulcanizator S. Finally, the as-prepared compounds were compression-molded at 150 °C, and the optimum curing time was obtained using a vulcameter. The resultant samples were denoted as illite-x/NR or C-illite-x/NR, with x representing the dosage of the illite or C-illite used, respectively.

Characterization

For structural characterization of CTAB-modified illite, FTIR spectra were recorded on a Bruker VERTEX 70 spectrometer (Germany) by averaging 32 scans at a 4 cm–1 resolution with the wavenumber ranging from 4000 to 400 cm–1. X-ray spectra were recorded using a Kratos Axis Supra X-ray photoelectron spectrometer (England) and normalized with the intensity of the Al2p spectrum as the internal reference. Wide-angle X-ray diffraction (WAXD) analysis was performed using a Xeuss 2.0 instrument (Xenocs, France) equipped with a Cu Kα radiation source (λ = 0.1542 nm).

Thermal gravimetric analysis (TGA) was measured on a STA8000F Frontier SQ8 (PerkinElmer, England) instrument under a nitrogen atmosphere at a heating rate of 10 °C/min.

The SEM experiments were performed using a scanning electron microscope (Phenom ProX, the Netherlands). The morphologies of pristine illite and tensile fractured surfaces of illite/NR and C-illite/NR composites were observed at the operating voltages of 15 and 10 kV, respectively. Before observation, the illite and fractured surfaces were sputter-coated with a thin gold film. Elemental analysis was characterized using energy-dispersive spectroscopy.

The vulcanization characteristics were observed at 150 °C on an M-3000A1 vulcameter (GOTECH, China). The cross-linking density was determined using the equilibrium swelling method and calculated using the Flory–Rehner equation.[56] The solvent used to determine the cross-linking density was toluene with a solubility parameter of 18.2, which is close to the solubility parameter of natural rubber (16.2–17.0). Three measurements were conducted for each sample, and the mean values with statistical errors are presented. The strain-dependent storage modulus (G′) of the rubber compounds and vulcanizates was analyzed using a rubber process analyzer (RPA2000, Alpha Technologies). The strain amplitude changed from 0.28 to 100% at the test frequency of 1 Hz and a temperature of 60 °C.

For mechanical property characterization, tensile and tear tests were performed using an INSTRON 5943 universal material testing machine at room temperature following ISO 37-2005 and ISO 34-1-2015 standards, respectively. Tensile tests were conducted on dumbbells of type 2 (ca. 75 mm × 4 mm × 2 mm) with a test length of 20 mm at an extension rate of 500 mm/min. Tear tests were carried out on angle test pieces (ca. 100 mm × 20 mm × 2 mm) with an extension rate of 500 mm/min. Five measurements for each sample were conducted to obtain an average value with the statistical error. The stress–strain curves presented show the most representative curve of the average. The Shore A hardness was tested according to the ISO 7619-1-2004 standard using the instrument GT-GS-MB from GOTECH Testing Machines Co. Five measurements at different positions on the test piece determine the median value of Shore A hardness. The abrasion resistance of vulcanizates was measured using a DIN abrader (GT-7012-D, GOTECH Testing Machines Co.) following the ISO 4649 standard. Three tests were conducted for each sample, and the mean value was chosen as the final result with the statistical error.