Improvement of mechanical properties of rubber compounds using waste rubber/virgin rubber.
For decades the rubber industry has been facing a major challenge to find a satisfactory way to deal with the enormous quantity of rubber goods (particularly tires) that reach the end of their life cycle. According to the End of Life Vehicle (ELV) Directive (2000/53/EC), carmakers must reuse or recover 85% of ELVs by weight per vehicle, and by the year 2006 at least 80% of that weight must be reused or recycled. For 2015 this target rises to 95% of ELVs by weight, 85% of which must be reused or recycled . Many attempts to recycle waste rubber have been made. The reuse of recycled rubber is important in terms of both waste disposal and reduction of product costs [2-5].
Waste rubber can be either ground into particles or devulcanized. According to the published results of numerous studies in the field of rubber recycling, ground rubber powder (GRP) and devulcanized rubber can be used in various products, including tires, profiles, construction articles, and mats [6-10]. When ground waste rubber was used as a component of different rubber products, a significant drop in the tensile strength was found, even at low levels of ground rubber . The deterioration of the mechanical properties , which limits the applicability of GRP in products, evidently occurs because the bonding force between the ground rubber and the matrix is often low, and thermodynamic incompatibility leads to poor interfacial adhesion. Devulcanized rubber was found to be suitable for use in some rubber compounds because the bonding force of the blends could be enhanced .
The objective of the present work was to study the effects of GRP and devulcanized rubber on the natural rubber (NR) matrix. The cure characterization, physical properties, and mechanical properties were examined. The effect of waste rubber on the mechanical properties, cross-link density, and cure characterization of virgin NR was investigated. It is possible to maintain the good mechanical properties of NR and obtain better mechanical properties, and at the same time lower the cost of the rubber material.
The virgin NR (SMR5) used as a matrix in this study was prepared according to ASTM D 3184 (formula 2A). SMR5 was masticated using 0.2-0.5 phr of a peptizing agent (Struktol A86; Schill & Seilecher). The GRP (about 4 mm), generated by ambient grinding of passenger car and light truck tires, was purchased from Genan A/S. It contained approximately 30%-wt NR, 40%-wt stirene-butadiene rubber (SBR), 20%-wt butadiene rubber (BR), and 10%-wt butyl- and halogenated butyl rubber (IIR/XIIR). The devulcanized rubber (supplied by Hauler Ltd., Finland) was produced from the above-mentioned GRP grade using a continuous shear flow reaction treatment under optimized process conditions. The processing temperatures were in the range of 180-230[degrees]C, and the screw speed was 375 rpm. The yield with these conditions was 80-100 kg/hr. Other chemicals used, such as zinc oxide, stearic acid, sulphur, N-220, and CBS, were of laboratory reagent (LR) grade.
The compound recipes used are given in Table 1. The loadings of the waste rubber (i.e., GRP and devulcanized rubber) were 10, 30, and 50 phr, respectively. The devulcanized rubber was first considered as an inactive additive (Dev.F), and in that case the amount of curatives used was based on the virgin rubber content. In another group of prepared samples (Dev.R), the amount of curatives used was based on the total rubber content, including virgin rubber and a raw rubber part from devulcanized rubber. In order to improve the mixing quality and prevent prevulcanization, a two-stage mixing process was used. In the first stage, the virgin NR was premixed at 60[degrees]C for 1 min in an internal mixer. After premixing, zinc oxide, stearic acid, carbon black, and waste rubber were gradually added to the mixer. The mixing continued for 5 min, after which CBS was added into the mixer for a further mixing of less than 1 min. In the second stage, an open two-roll mixing mill was used at 60[degrees]C for 1 min to soften the premixture, after which sulphur was added. The mixing was continued for another 2.5 min and then the mixture was cut into sheets, which were cooled to room temperature. The sheets were molded at 160[degrees]C in an electrically heated compression press. The coarse structure of the GRP caused some roughness on the surface of the molded 2-mm-thick sheets.
A Monsanto 100S oscillating disc rheometer was used to obtain the cure characteristics at 160[degrees]C according to ISO 3417.
Swelling Ratio and Cross-Link Density
The cross-link densities of the blends were obtained from swelling experiments in toluene at room temperature. Circular test pieces with a diameter of 10 mm were cut from the vulcanized sheets and dried overnight in a vacuum desiccator. The specimens were accurately weighed ([M.sub.i]) and immersed in toluene in closed bottles for 3 days. Then the surfaces were dried with filter paper and the samples were quickly weighed ([M.sub.t] is the swollen weight of the vulcanizates after equilibrium is reached). The rubber swelling ratio Q in solvent was calculated using the equation:
Q = [[M.sub.t] - [M.sub.i]]/[M.sub.i]. (1)
For vulcanizates that contain reinforcing fillers, such as carbon black, the volume fraction of rubber [v.sub.r] for use in the calculation of cross-link density is obtained from the following expression derived by Kraus  and Sobhy et al. :
[v.sub.r]/[v.sub.rf] = 1 - m[phi]/(1 - [phi]) (2)
where [v.sub.rf] and [v.sub.r] are the volume fractions of filled and unfilled rubber in the swollen rubber phase, respectively; m is the rubber-filler interaction parameter; and [phi] is the volume fraction of the filler.
The cross-link density of the gels was calculated using the Flory-Rehner equation [16, 17]:
v = [ln(1 - [v.sub.r]) + [v.sub.r] + [chi][v.sub.r.sup.2]]/[2[V.sub.s](0, 5[v.sub.r] - [v.sub.r.sup.1/3])] (3)
where v is the cross-link density, [V.sub.s] is the molar volume of the swelling solvent, [v.sub.r] is the rubber fraction from the Kraus equation, and [chi] is the polymer solvent interaction parameter ([chi] = 0.393 for the NR-toluene system) .
The tensile and tear properties were measured using a Lloyd testing machine. The tensile specimens were dumbbell-shaped test pieces according to ISO 37 (type 1), and the tear resistance was determined using trouser test pieces according to ISO 34 (type A). The speed in the tensile tests was 500 mm/min and 100 mm/min for the tear strength. Abrasion resistance was carried out according to ISO 4649 using an abrasive run of 40 m and loading 10 N. Hardness was tested based on ISO 7619 using a ShoreA durometer.
[FIGURE 1 OMITTED]
The morphology of the tensile fractured surfaces of the samples was investigated using a scanning electron microscope (JEOL JSM-T100). The fractured surfaces were sputter-coated with gold powder using a K550 sputter coater.
RESULTS AND DISCUSSION
The effects of waste rubber loading on the cure characteristics are shown in Figs. 1 and 2.
[FIGURE 2 OMITTED]
Figure 1 shows that the minimum torque ([M.sub.L]) changed slightly with the increase of devulcanized rubber content, but rose significantly when the GRP content was increased. This indicates that it is more difficult to process compounds that contain GRP compared to compounds with devulcanizea rubber. The apparent reason for the minimum torque increase is that cross-linked GRP does not easily flow to the matrix, so the increase in GRP loading may reduce the flow and consequently increase the torque. It is also believed that the highly aggregated and convoluted structure of waste rubber powder contains void space in which the matrix rubber is trapped, which increases the effective volume fraction of waste rubber and results in higher viscosity of the blends.
The maximum torque ([M.sub.H]), which is a measure of the elastic modulus, was lower with increased Dev.F loading, while it remained stable in the GRP systems. It is important to observe the increase in [M.sub.H] with the addition of devulcanized rubber when it is considered as a part of rubber. This indicates that the elastic modulus is lower in the compounds with Dev.F content. The shortest fragments and smaller chains in the devulcanized rubber will act as plasticizers, decreasing the viscosity and torque of the compounds and increasing their tack. The higher elastic modulus in GRP compounds led to a greater maximum torque in all three GRP systems, in which a dramatic increase in [M.sub.H] was clearly observed with the introduction of Dev.R. The torque increments with Dev.R compounds could be explained by further revulcanization and cross-link reaction with the devulcanized rubber, due to the new active sites formed during the devulcanization process and to the amount of curatives in Dev.R compared to the Dev.F system.
[FIGURE 3 OMITTED]
The scorch time ([t.sub.s2]) and optimum cure time ([t.sub.90]) of the compounds did not change significantly (see Fig. 2). The scorch time decreased in all systems, with the GRP system having the lowest and the Dev.F system the highest scorch time values. The shorter scorch time in the blends indicates that the cross-linking reaction started earlier. A product of a reaction between the NR matrix and the devulcanized rubber may catalyze the cross-linking reaction . Also, diffusion of the accelerator from devulcanized rubber into virgin rubber would reduce scorch time . Since the scorch time slightly decreased in all blends, the processibility did not change significantly with the waste rubber loading.
The optimum cure time is the vulcanization time required to obtain the optimum physical properties. It can be observed in Fig. 2 that the optimum cure time varies slightly between the GRP and Dev.F systems. The cure time is shorter in the Dev.R system because it contains more curatives than the other two systems, and the devulcanized rubber participates together with virgin rubber in the cross-linking reaction. In summary, it can be concluded that the incorporation of waste rubber into the virgin NR matrix does not effect the cure time significantly.
The cure rate index (CRI) , which is a measure of the rate of the cure reaction, is given by:
CRI = 100/([t.sub.90] - [t.sub.S2]). (4)
It is evident from Fig. 3 that the compounds acquire a lower cure rate with the introduction of GRP. The devulcanized rubber participates in the cross-link reaction as a part of rubber when a sufficient amount of curative is present. For this reason the cure rate remains the same as for virgin rubber in the Dev.R system. On the other hand, the possibility of cross-link formation between unsaturated sites of devulcanized rubber and virgin rubber has been proposed. The lower CRI in Dev.F compounds may be due to the lack of curing agents. It is possible that the GRP of such a cross-linked network hinders the mobility of the unsaturated segments of NR, thus accounting for the lowest CRI for this system.
Swelling Ratio and Cross-Link Density
Figure 4 shows the swelling ratio of NR compounds of various waste rubber contents in toluene. The swelling ratio increased with the GRP loading in NR matrix, as shown in Fig. 4. This is attributed to the difficulty of diffusing sulphur in the rubber matrix because of obstruction by ground vulcanizates . As for the compounds containing devulcanized rubber, it is evident that the GRP is partially devulcanized and partially de-polymerized during devulcanization. The swelling ratio also increased with Dev.F content because of shorter molecules and smaller fragments in the devulcanized rubber. However, there was a decrease in the swelling ratio in the Dev.R compounds, in which the devulcanized rubber was assumed as a part of rubber. This is only to be expected when one considers the possibility for further cross-link reactions in both devulcanized rubber and virgin rubber with a sufficient amount of curatives.
Figure 5 presents the variations in cross-link density of the systems as a function of waste rubber content. It can be observed that the cross-link density decreased with an increase in GRP and Dev.F loading. This is due to the migration of sulphur being obstructed by the waste rubber particles. There are new active cross-linking sites in devulcanized rubber that continue to form a cross-link network during revulcanization. The more Dev.R, the more the active cross-linking sites. Hence, the cross-link density is improved in the Dev.R system. Higher cross-link density leads to improved mechanical properties, which also explains the higher maximum torque of Dev.R compounds.
[FIGURE 4 OMITTED]
Tensile strength, elongation at break, tear strength, abrasion resistance, and hardness were used to evaluate the mechanical properties of the systems. The values are listed in Table 2. Tensile strength and elongation at break were impaired with the loading of both GRP and devulcanized rubber. The reason is probably the uncontinuous and imperfect structure of the blends. However, the Dev.R compounds have much better properties compared to the Dev.F compounds. This improvement in mechanical properties indicates a higher degree of cross-linking in Dev.R. As could be expected, the compounds that contained GRP showed the lowest tensile strength and the lowest elongation at break. However, better compatibility and migration between the devulcanized rubber and the NR matrix were achieved and better properties were obtained (see Table 2). It is worth noting that with up to 10 phr loading of Dev.R, the compound had even higher tensile strength and the same elongation as virgin rubber, and it retained 89% of its original strength and 60% of its elongation at break with 50 phr Dev.R loading.
[FIGURE 5 OMITTED]
The effect of the waste rubber content on tear strength is given in Table 2. The addition of waste rubber components to the virgin rubber matrix did not produce any significant effect on the tear properties of the compounds. However, at the highest concentration of both GRP and devulcanized rubber, the tear strength showed some improvement compared to the virgin NR compound. This result indicates rather good adhesion between the matrix and the added components.
The variation in abrasion resistance with waste rubber content is shown in Table 2. The results indicate that abrasion resistance is improved with the GRP and Dev.R loading, but impaired with the Dev.F loading. The higher modulus with vulcanized particles of GRP leads to better abrasion resistance for the GRP compounds. The abrasion resistance is impaired because of the shortest fragments and smaller chains in Dev.F: the greater the Dev.F loading, the worse the abrasion resistance. However, an important increase in abrasion resistance with Dev.R loading was observed because revulcanization and more cross-linked networks occurred in the Dev.R system, and higher cross-link density leads to better abrasion resistance.
[FIGURE 6 OMITTED]
The decrease in hardness in the compounds was slight with an increase in Dev.F loading (see Table 2). The value for 10 phr Dev.F was the same as for the NR matrix, while the hardness with 50 phr Dev.R loading retained 93% of the virgin NR. The shortest fragments and lowest network density in the Dev.F decreased the hardness and impaired the mechanical properties. In comparison, the systems with GRP and Dev.R had better hardness because of the higher modulus and cross-link density in the compounds. Again, GRP rubber is probably harder than virgin NR.
[FIGURE 7 OMITTED]
As can be seen in Table 2, the standard deviations (SDs) in the test results were considerably higher for the samples that contained waste rubber components compared to the NR samples. The lowest SD values for the waste rubber compounds were shown by the Dev.R materials. The highest deviations in the results can be explained by the inhomogeneous structure of the materials.
[FIGURE 8 OMITTED]
SEM photomicrographs of the tensile fractured surface of three compounds are shown in Figs. 6-8. In Fig. 6, the fractured surface of compound GRP at 10 phr loading reveals that the phase separation is more clearly observable in the GRP system. Phase separation of the Dev.F system is also observable in Fig. 7. The Dev.R system has much better compatibility and the phase separation becomes less marked (Fig. 8), which is in agreement with the revulcanization and further cross-link reaction between Dev.R and the NR matrix.
Waste rubber of 10, 30, and 50 phr was added to a virgin rubber matrix in the form of GRP and devulcanized rubber. The same GRP grade that was used as the starting material in the devulcanizing process was applied in the GRP compounds.
It was found that the addition of devulcanized rubber together with an adequate amount of curatives (Dev.R) produced the best mechanical performance. With the exception of abrasion resistance, the mechanical properties of the compounds that contained GRP were impaired in comparison to the other waste rubber and virgin NR compounds. The mechanical properties of the waste rubber compounds can be set in the following ranking order: Dev.R > Dev.F > GRP. This observation is supported by the higher cross-link density and SEM tensile fractured surface in the Dev.R system, which show that Dev.R compounds have the best compatibility and GRP compounds the worst compatibility in the case of waste rubber in a virgin rubber matrix.
1. ELV Directive, Official Journal of the European Communities, L269, p 34 (21, 10, 2000).
2. K. Nasker, A.K. Bhowmick, and S.K. De, Polym. Eng. Sci., 41, 1087 (2001).
3. E. Sipahi-Saglam, C. Kaynak, G. Akovali, M. Yetmez, and N. Akkas, Polym. Eng. Sci., 41, 514 (2001).
4. R. Tripathy, J.E. Morin, D.E. Williams, S. Eyles, and R.J. Farris, Macromolecules, 35, 4616 (2002).
5. J. Pfretzschner and R.M. Rodriguez, Polym. Test., 18, 81 (1999).
6. C. Jacob, A.K. Bhowmick, P.P. De, and S.K. De, Plast. Rubber Compos., 31, 212 (2002).
7. W. Klingensmith and K. Baranwal, Rubber World, 218, 41 (1998).
8. M.C. Bignozzi, A. Saccani, and F. Sandrolini, Composites: Part A, 33, 205 (2002).
9. F. Padella, F. Cavalieri, G. D'Uva, A. La Barbera, and F. Cataldo, Polym. Recycl., 6, 11 (2001).
10. N. Sunthonpagasit and M.R. Duffey, Resour. Conserv. Recycl., 40, 281 (2004).
11. J.K. Kim and R.P. Burford, Rubber Chem. Technol., 71, 1028 (1998).
12. H. Ismail, R. Nordin, and A. Md. Noor, Polym. Plast. Technol. Eng., 41, 847 (2002).
13. M.J. Myhre and D.A. MacKillop, Rubber World, 214, 42 (1996).
14. G. Kraus, J. Appl. Polym. Sci., 7, 861 (1963).
15. M.S. Sobhy, D.E. El-Nashar, and N.A. Maziad, Egypt. J. Sol., 26, 241 (2003).
16. M.A.L. Verbruggen, L. Van Der Does, and J.W.M. Noordermeer, Rubber Chem. Technol., 72, 731 (1999).
17. C. Kumnuantip and N. Sombatsompop, Mater. Lett., 57, 3167 (2003).
18. B.G.C.J. Wijers, in Proceedings of the International Rubber Conference (IRC), Birmingham, UK, 380 (2001).
19. U.S. Ishiaku, C.S. Chong, and H. Ismail, Polym. Polym. Compos., 6, 399 (1998).
20. C.K. Hong and A.I. Isayev, J. Mater. Sci., 37, 385, (2002).
21. A.R.R. Menon, C.K.S. Pillai, and G.B. Nando, Polymer, 39, 4033 (1998).
22. H. Ismail, R. Nordin, and A.M. Noor, Polym. Test., 21, 565, (2002).
Shuyan Li, Johanna Lamminmaki, Kalle Hanhi
Plastics and Elastomer Laboratory, Institute of Materials Science, Tampere University of Technology, P.O. Box 589, 33101 Tampere, Finland
Correspondence to: S. Li; e-mail: Shuyan.firstname.lastname@example.org
TABLE 1. The compound recipes. Formulation (phr) NR Devulcanized rubber GRP (4-mm) Zinc oxide NR 100 -- -- 5 GRP 10 (b) 100 -- 10 5 30 100 -- 30 5 50 100 -- 50 5 Dev. F 10 (c) 100 10 -- 5 30 100 30 -- 5 50 100 50 -- 5 Dev. R 10 (d) 96.6 3.4 -- 5 30 89.8 10.2 -- 5 50 83 17 -- 5 Formulation (phr) Stearic acid N-220 CBS (a) Sulphur NR 2 35 0.7 2.25 GRP 10 (b) 2 35 0.7 2.25 30 2 35 0.7 2.25 50 2 35 0.7 2.25 Dev. F 10 (c) 2 35 0.7 2.25 30 2 35 0.7 2.25 50 2 35 0.7 2.25 Dev. R 10 (d) 2 35 0.7 2.25 30 2 35 0.7 2.25 50 2 35 0.7 2.25 (a) N-cyclohexyl-2-benzothiazyl sulphonamide, as an accelerator. (b) Amount of curatives was based on the virgin rubber content. (c) Amount of curatives was based on the virgin rubber content. (d) Amount of curatives was based on the total rubber content, including virgin rubber and devulcanized rubber. TABLE 2. Mechanical properties of the compounds. Formulation Tensile strength Elongation at break (phr) (Mpa) SD (a) (%) SD NR 23.2 0.5 719.5 9.9 GRP 10 (b) 13.1 4.2 256.2 76.5 30 9.6 2.1 313.8 43.3 50 6.5 1.0 241.0 18.2 Dev. F 10 (c) 20.4 1.9 438.5 106.3 30 14.6 2.9 379.1 51.2 50 13.4 3.3 422.7 64.6 Dev. R 10 (d) 26.3 0.5 522.8 12.9 30 20.4 1.8 469.0 26.0 50 20.7 0.7 425.8 11.8 Formulation Tear strength Abrasion Hardness (phr) (N/mm) SD ([mm.sup.3]) SD (ShoreA) SD NR 20.8 2.2 150 7 54.0 0.5 GRP 10 (b) 20.1 6.9 111 16 59.5 0.5 30 16.8 3.6 120 1 61.0 6.0 50 22.5 6.9 132 11 59.5 0.5 Dev. F 10 (c) 20.6 8.3 157 8 54.0 1.5 30 19.9 2.8 210 10 52.5 1.5 50 27.0 2.9 239 3 51.0 0.5 Dev. R 10 (d) 17.1 1.9 122 4 57.5 1.5 30 19.7 2.4 132 1 60.0 0.5 50 22.4 4.7 132 1 64.5 0.5 (a) Standard deviation. (b) Amount of curatives was based on the virgin rubber content. (c) Amount of curatives was based on the virgin rubber content. (d) Amount of curatives was based on the total rubber content, including virgin rubber and devulcanized rubber.
|Printer friendly Cite/link Email Feedback|
|Author:||Li, Shuyan; Lamminmaki, Johanna; Hanhi, Kalle|
|Publication:||Polymer Engineering and Science|
|Date:||Sep 1, 2005|
|Previous Article:||Rheological behavior comparison between PET/HDPE and PC/HDPE microfibrillar blends.|
|Next Article:||Influence of annealing treatment on the heat distortion temperature of nylon-6/montmorillonite nanocomposites.|