Eco-plastics: morphological and mechanical properties of recycled poly(carbonate)-crushed rubber (rPC-CR) blends.
Because of both economical and environmental reasons, there is an important need to find a second life use for all wastes. Among all end of life rubber goods, used tires constitute one of the largest source of rubber wastes. Although rubbers are quite impossible to recycle (at least physically), because they are crosslinked polymers, they can be ground and used as more or less functional fillers in a wide variety of polymer matrices as recently reviewed by De [1, 2]. In fact, finely crushed rubber (CR) wastes obtained by several milling techniques can be dispersed in new formulations of rubbers [3-6], thermoplastics [7-13], thermoset resins [14, 15] or asphalt .
In most cases, physical properties and processability are reported to be affected when large amounts of CR particles are added to a polymer matrix. To obtain composites with acceptable properties, the quality of the interface between CR particles and the polymer matrix needs to be improved. This can be achieved using physical and chemical phenomena. In the first case, a decrease of particle size by grinding together with a solvent etching will increase the specific surface of the filler and thus the interface size. As a consequence, the number of possible interactions at the interface will be increased. In the second case, the intensity of interactions can be enhanced grafting the filler surface with chemical functions able to either interact or even react with the macromolecular chains of the matrix. Several examples of filler/matrix compatibilization can be found in the literature, from simple surface treatment of fillers , coupling agent addition [16, 17], to reactive blending [18-22].
When a good adhesion between the fillers and the matrix is achieved positive synergies can be reached as impact strength increase  toughness increase  and elongation at break increase . Rubber particles are expected in one hand to stop crazes helping to dissipate energy and on the other hand to decrease the composite's Young modulus. It is known that neat PC like some other unmodified plastics show severe notch sensitivity and may fail in a brittle manner by crazing, which is initiated essentially by the high hydrostatic stresses prevailing at notches and crack tips . Blending small rubber particles of submicron size with a PC matrix leads to cavitation of these rubber particles, which facilitates plastic "shear yielding" in the surrounding matrix and at the same time delays or suppresses crazing by the relief of hydrostatic stress.
This aim of this study is to show the opportunity to develop formulations of recycled poly(carbonate) (rPC) and CR particles to obtain eco-friendly plastics (EFP) with satisfying mechanical properties. We have first studied the influence of filler content, particles size, rubber surface treatment, and compatibilization of fillers with the matrix, on mechanical characteristics. In parallel, the results have been interpreted in terms of composites morphology and filler/matrix interactions.
Production wastes of Bayloy[R] (AXXIS) poly(carbonate) sheets were obtained from self-signal society. This formulation is based on Makrolon[R] 3103 (BAYER). The main characteristics of both virgin poly(carbonate) (vPC) and rPC are summarized in Table 1. Additional thermo-mechanical data on rPC can be found in reference  in which the evolution of thermal and mechanical properties of PC is followed over eight successive recyclings. Different sizes of raw CR particles were obtained from a recycler, Delta-Gom. A previous study has shown that as received CR could not lead to satisfying mechanical properties . Hence, it was necessary to grind them into a mill containing liquid nitrogen to reduce their size. To obtain several ranges of diameter ([phi]) particles were sieved: 140 < [phi] < 315 [micro]m and [phi] < 140 [micro]m. Fine micro-metric CR particles were then melt mixed with millimet-ric PC crushed pellets. CR density measured with a pyc-nometre was d = 0.840. Two types of surface treatments were carried out: first, a flame treatment proceeded to oxidize the CR surface and obtain satisfactory level of adhesion with poly(carbonate). This treatment was performed with a propane blowtorch. The flame temperature was about 800[degrees]C and particles were flamed during 1 min at a distance of 20 cm. Second, a solvent washing with dichloromethane was done to eliminate oils residues. Particles were dispersed in solution under sonication and stirring at room temperature for 25 min. Then the particles were filtered and dried under vacuum at 40[degrees]C for 30 min. Two types of compatibilizers were used: a triblock copolymer of ethylene, ethylacrylate, and hydroxyl methacrylate (E-EA-MAH) referenced as LOTADER AX4700 and a triblock copolymer of ethylene, methylacrylate, and glycidyl methacrylate (E-MA-GMA) referenced as Lotader AX8920. The main mechanical and physical properties of the compatibilizers are listed in Table 2.
Melt Mixing. PC/CR blends were melt mixed in a Bra-bender 50 EHT internal mixer with contra rotating blades driven by WinMix software. Polymers were dried under vacuum for 24 h at 90[degrees]C before processing. PC matrix and CR particles were mixed with a rotor speed of [ohm] = 40 rpm at T = 240[degrees]C for 10 min. These optimized blending conditions allowed a good dispersion of CR into PC. Different formulations were studied (all contents are in wt%) in this work: 100% rPC, 95%rPC/5% untreated CR, 90% PC/10% untreated CR, 80% PC/20% untreated CR, 80% PC/20% flame treated CR, 80% rPC/20% solvent treated CR, 77% rPC/20% CR/3%AX4700, and 77% rPC/20% CR/3%AX8920. For mechanical testing specimens were processed in two steps. A first compression molding (at [T.sub.mold.sup.[dot]] = 240[degrees]C, [p.sub.mold.sup.[dot]] = 50 bar, [t.sub.mold.sup.[dot]] = 5 min) allows obtaining larges plates of about 4 x 250 x 250 [mm.sup.3]. In the second step, after the plates had been brought down to room temperature in ~15 min, normalized samples was obtained using a small numerical milling machine.
Morphology Analysis by Microscopy. Scanning electronic microscopy (SEM) observations to determine either filler or composites morphologies were performed with a JEOL JSM-6031 after fracture of the sample in liquid nitrogen and spray deposition onto the surface of a thin gold layer.
Particle dimensions were measured with a Leica MZ3 optical microscope with Lida software in diascopy mode and non polarized light.
Mechanical Testing. Tensile tests were carried out on a MTS Synergie RT1000 machine, using standard IS0527 samples with a crosshead speed of 1 mm [min.sup.-1]. Five tensile specimens were tested for each formulation. Indentation tests were performed with a commercial XP Nano-indenter from MTS Nano Instruments at room temperature (23 [+ or -] 1[degrees]C) with a continuous stiffness measurement (CSM) technique. A three-side pyramid (Berkovitch) diamond indenter was used for the indentation tests. Additional details concerning the experimental protocol can be found in Ref. 26.
RESULTS AND DISCUSSION
Both morphologies of the blends and their mechanical properties have been studied in parallel for the different formulations. In each case tension tests results are interpreted to the light of blends morphologies analysis and in terms of filler matrix interactions.
Effect of CR Particles Size, Shape and Content
Morphology of Composites. Several kinds of morphologies have been observed: size and shape of filler, dispersion of fillers, fillers surface, and filler/matrix adhesion.
From Fig. 1, it can be seen that as received CR particles are very different in shape and size. The rather geometric shape of CR particles resulting from their two successive crushing is difficult to describe precisely. Nevertheless, two populations of CR particles diameter ([phi]) have been obtained using sieves of normalized mesh: 140 < [phi] < 315 [micro]m and [phi] < 140 [micro]m. In the following these two ranges of diameter will be named respectively large and small particles. The mean dimensions of particles were checked by microscopic measurements and the diameter partitions were confirmed.
Comparing Figs. 2 and 3 illustrates the simple fact that dividing the filler results in an increase of the interface size and particle number. It can also be seen that adhesion between CR and rPC is not improved as in both cases a detachment of particles from the matrix is clearly visible. Nevertheless, the contact surface increase due to the higher level of fractionation of fillers is expected to enhance the work of adhesion between CR and rPC. Moreover, CR particles are found to be well dispersed and not to tend to aggregate themselves.
[FIGURE 1 OMITTED]
[FIGURE 2 OMITTED]
Mechanical Properties. Figures 4 and 5 clearly show that both tensile strength and Young modulus of rPC/CR blends decrease gradually with increasing CR content and this much quicker with large particles. This is well illustrated in Fig. 4 by the fact that for the same particles content, the tensile strength obtained with the small particles is about 30% more important. This phenomenon is less sensitive with modulus decrease where no important difference is found due to the particle size up to 15% CR. The influence of particles size is certainly even more visible on strain at break in Fig. 6. Large particles increase dramatically the blend brittleness, whereas small particles have quite no effect up to 25% of CR. These important changes of mechanical properties obtained just by decreasing the rubber particles size under 140 [micro]m and despite poor adhesion between rPC and CR can attribute the interface size increase, to a higher level of mechanical anchorage of PC matrix at particles surface and also to better dispersion of particles within the matrix during blending. The fact that whatever particles size all samples gave rise to a fragile behaviour under tensile test can be explained by the poor adhesion between the two polymer phases but also because of poly(carbonate) first recycling, which already decreased its molar mass . Consequently, inelastic deformation behaviour usually displayed by amorphous thermoplastic polymers does not take place and the little hardening observed in [24, 27] of the material due to a better alignment of the macromolecular chains became shorter during processing, does not occur. Thus poly(carbonate) becomes less rigid and brittle compared with its virgin state as showed in Table 1. Knowing the effect of particle size, we have investigated the influence of the enhancement of interaction level between CR particles and rPC matrix on mechanical properties of rPC/CR blends. In the following, mechanical testing has been limited to blends containing large particles and morphological analysis to blends filled with 20% by weight of CR of both kinds of particles.
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
Effect of CR Surface Treatments
Two types of particles surface treatment have been used to improve mechanical properties, flaming, and washing (protocols are described in the "techniques" paragraph).
Morphology of Composites. Comparing Fig. 7 with Fig. 3 shows that the morphologies of rPC/CR flamed blends are rather different from that of rPC/CR untreated blends. This time one can notice a better wetting of CR by rPC matrix and a weak detachment of CR particles. CR size does not seam to have any importance on the interface cohesion. At the contrary, comparing Fig. 8 shows that a better interfacial adhesion is suspected for small particles washed by dichloromethane.
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
Mechanical Properties. The main feature coming from Fig. 9 is that rather surprise washing is the most effective treatment to slowdown the tensile strength decrease with CR content increase. On the other hand, Figs. 10 and 11 show that flaming is the most effective treatment to stabilize both Young's modulus and strain at break with CR content. Even if washing is not the most effective treatment to stop the modulus decrease, it prevents the catastrophic drop over 16% w/w of CR. But, how to interpret the effect of these two different surface treatments on the composite mechanical properties? For dichloromethane washing, the adhesion increase results from elimination from the surface of oils residues and pollution as evidenced by chromatographic analysis of the washing solution. For flaming treatment,, adhesion enhancement can be attributed to the apparition of polar functions at the rubber surface able to interact with dipoles on rPC chains. But to go further in the interpretation of these phenomena, it is worth looking at nanoindentation results obtained with the same composites (cf. Fig. 12) and more extensively studied in another study . One can see in this figure the evolution of the modulus decrease through the rPC/CR interphase. To highlight the modulus decrease, data measured across the interphase have been normalized by subtraction of rPC/CR untreated moduli at the same indent position, which explains the negative ordinate scale. Washing appears to strongly modify the interphase modulus profile. The decrease of rPC and CR moduli observed can be explained by the diffusion of rubber free chains out of the cross-linked rubber network into the rPC matrix during the blending. This diffusion process could in one hand slightly plastisize rPC matrix in the vicinity of CR particles and also soften CR network. Such deep modification of the interphase can explain the increase of tenacity enhancement resulting from the washing treatment. It can be noticed that the modulus drop is only shifted of 2 [micro]m from the frontier between the two materials determined by optical microscopy. In comparison with the flaming treatment, one can see that the modulus break appears at about 5 [micro]m from the interface, which suggests a different mechanism of adhesion. It is most likely that flaming limits the diffusion of small chains out of rubber into rPC and additionally improves interactions between the two polymers thus explaining the modulus and strain at break increase.
[FIGURE 9 OMITTED]
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
[FIGURE 12 OMITTED]
In all cases, mechanical properties of blends containing untreated large CR are clearly lower, indicating that both treatments induce significant increase of adhesion at the interface. As a result, it is possible using rather simple treatments (washing and flaming) to slow down the mechanical properties decreased expected from addition of large amount of CR micro particles into rPC and this up to 24% of CR w/w. Moreover, data collected for composites containing 20% w/w of CR in Table 3 show the impact of both particle size and surface treatment on mechanical properties, i.e., strength at break, Young's modulus, and strain at break. It can be noticed that in some cases it is more useful to decrease the particle size than to do any surface treatment of CR, as for example, stress and strain at break (comparing Fig. 4 with Fig. 9 and Fig. 5 with Fig. 10). Nevertheless, flaming allows obtaining the highest modulus.
[FIGURE 13 OMITTED]
Effect of Compatibilizer Addition
Morphology of Composites. The Figs. 13 and 14 show a better wetting of particles surface by the rPC matrix and a higher homogeneity of the fracture surface due to addition of compatibilizers. This suggests a better dispersion of fillers and a decrease of particle coalescence. Consequently, a decrease of interfacial tension is expected and thus a more important adhesion between micro particles of CR and rPC matrix. The addition of compatibilizers should also lead to more regular and stable morphologies compared with untreated, flamed, and washed CR. We will now study the influence of the compatibilizer's chemical nature on the composites mechanical properties.
[FIGURE 14 OMITTED]
[FIGURE 15 OMITTED]
Mechanical Properties. The graphs of Figs. 15 and 16 give an overview of mechanical characteristics of rPC-20% w/w CR (large particles) composites obtained respectively in elastic domain and at break. First of all, it is clear that all treatments increase both stress and strain compared with untreated CR composites, unlike for modulus, which decreases in all cases resulting from a compa-tibilization effect, excepted for flaming.
[FIGURE 16 OMITTED]
In the linear domain of strain, the surface treatments are globally more effective to increase stress and strain of composites than the coupling agents. In the break zone, this tendency is maintained for stress and reversed for strain. To go further, one can see that the mechanical behaviours resulting from the two types of treatments are radically different: the surface treatments tend to extend the elastic zone without increasing the plastic zone, whereas the main effect of coupling agents is to increase the plastic zone and slightly the elastic domain together with a weak decrease of the Young's modulus. Thus depending on the target application, it is possible to adjust the recycled composites mechanical properties changing the nature of micro-particles treatment. Nevertheless, it must be kept in mind that whatever the treatment an eco-plastic obtained dispersing 20% w/w of CR into rPC will not reach properties higher than that of pure poly(carbonate). But the properties obtained may be sufficient for not too demanding applications, allowing to decrease costs and impact on environment together with a preservation of nonrenewable feedstocks.
Because of the large number of applications requiring the use of poly(carbonate) and reticulated rubbers, there are important quantities of wastes of end of life products made of these polymers. Especially the later family of polymers, which cannot be melted can hardly find a new life in the same kind of application. This is why we have chosen to associate rPC from signalization panel to CR from car tires and study the mechanical properties of the resulting composite to develop new EFPs with low impact on environment.
In the first step, we have found that an important gain of mechanical properties could be obtained using small particle, i.e., with size < 140 [micro]m. Nevertheless, as the grinding and sewing processes necessary to obtain such small rubber particles are complex and expansive, it was decided to use another route to achieve the same level of mechanical properties. Thus two kinds of particle treatments were experimented, surface activation by either washing and flaming and compatibilization with coupling agents. The first treatment was found to extend significantly the elastic domain of the EFP even at high CR contents of about 20% w/w. The action of the coupling agents on the EFP mechanical properties was mostly to complement the elastic domain with a more important plastic zone. Almost all treatments were found to enhance stress and strain but to decrease the Young's modulus. In best conditions, the particle treatment allowed to maintain the level of properties of poly(carbonate) with increasing CR content up to 25% w/w) but not to provide significant enhancement of mechanical properties. The choice of suitable particle treatment and may be their combination should allow to tailor the EFP mechanical properties to reach a sufficient level for not too demanding applications.
Finally, the very promising environmental gains, which should result from the use of such EFP obtained from full recycled polymers seems to be one of the most important contributions of the present research.
The authors thank F. Boulay, J. Babois, H. Bellegou, H. Guezenoc, and F. Peresse for their contribution to this work, DeltaGom and Self Signals for supply of respectively crushed rubber particles and recycled poly(carbonate).
1. S.K. De, Progr. Rubber Plast. Recycl. Technol., 17(2), 113 (2001).
2. B. Adhikari, D. De, and S. Maiti, Prog. Polym. Sci., 25, 909 (2000).
3. A.A. Phadke, S.K. Chakraborty, and S.K. De, Rubber Chem. Technol., 57(1), 19 (1984).
4. A.A. Phadke, A.K. Bhowmick, and S.K. De, J Appl. Polym. Sci., 32, 4063 (1986).
5. D. Gibala, D. Thomas, and G.R. Hamed, Rubber Chem. Technol., 72, 357 (1999).
6. A.K. Naskar, P.K. Pramanik, R. Mukhopadhyay, S.K. De, and A.K. Bhowmick, Rubber Chem. Technol., 73(5), 902 (2000).
7. A.A. Phadke and S.K. De, Polym. Eng. Sci., 26, 1079 (1986).
8. P. Rajalingam, J. Sharpe, and W.E. Baker, Rubber Chem. Technol., 66, 664 (1993).
9. K. Oliphant and W.E. Baker, Polym. Eng. Sci., 33, 166 (1993).
10. P.K. Pramanik and W.E. Baker, Plast. Rubber Compos. Process. Appl., 24, 229 (1995).
11. T. Luo and A.I. Isayev, J. Elast. Plast., 30, 133 (1998).
12. A.K. Naskar, A.K. Bhowmick, and S.K. De, Polym. Eng. Sci., 41, 1087 (2001).
13. A.K. Naskar, D.P. Gala, S.K. De, and A.K. Bhowmick, Kautsch. Gummi Kunstst., 4, 164 (2002).
14. R. Bagheri, M.A. Williams, and R.A. Pearson, Polym. Eng. Sci., 37, 245 (1997).
15. X. Colom, F. Carillo, and J. Canavate, Compos. A, 38(1), 44 (2007).
16. P. Rajalingam and W.E. Baker, Rubber Chem. Technol., 65, 908 (1992).
17. S. Datta and S.D. J. Lohse, Polymeric Compatibilizers Uses and Benefits in Polymer Blends, Hanser, Munich (1996).
18. H. Scholz, P. Potschke, H. Michael, and G. Mennig, Kautsch. Gummi Kunstst., 55(11), 584 (2002).
19. C. Radhesh Kumar, I. Fuhrmann, and J. Karger-Kocsis, Polym. Degrad. Stab., 76, 137 (2002).
20. C. Radhesh kumar and J. Karger-Kocsis, J. Plast. Rubber Compos., 31(3), 99 (2002).
21. S. Wiessner, H. Michael, and G. Mennig, Kautsch. Gummi Kunstst., 56(10), 514 (2003).
22. S. Wiessner, U. Wagenknecht, M. Zichner, H. Michael, and G. Heinrich, in 21st Polym. Proc. Soc., Leipzig, SL 4-5, June 19-23 (2005).
23. A.M. Donald, "Crazing", in The Physics of Glassy Polymers, Chapman and Hall, London, 295-341, (1997).
24. J.F. Feller and A. Bourmaud, Polym. Degrad. Stab., 82, 99 (2003).
25. K. Zribi, J.F. Feller, A. Bourmaud, K. Elleuch, and B. Elleuch, Polym. Adv. Technol., 17(9/10), 727 (2006).
26. A. Bourmaud, J.F. Feller, Y. Grohens, and K. Zribi, J Appl. Polym. Sci., 103(4), 2687 (2006).
27. M.C. Boyce, D.M. Parks, and A.S. Argon, Mech. Mater., 7, 15 (1988).
K. Zribi, (1,2) K. Elleuch, (3) J.F. Feller, (1) A. Bourmaud, (1) B. Elleuch (2)
(1) Laboratory of Polymers, Properties at Interfaces and Composites (L2PIC), University of South Brittany, Lorient, France
(2) Laboratory of Water, Energy, and Environment, National School of Engineers, Sfax, Tunisia
(3) Laboratory of Electromechanical System, National School of Engineers, Sfax, Tunisia
Correspondence to: Jean-Francois Feller; e-mail: email@example.com
Contract grant sponsor: French Embassy, Tunisia (French Ministry of Foreign Affairs); French Ministry of Research and New Technologies and the Brittany Region.
TABLE 1. Virgin and recycled poly(carbonate)s characteristics, respectively, vPC and rPC. vPC rPC [T.sub.g] ([degrees]C), glass transition temperature 148 149 E (GPa), Young's modulus 2.3 1.95 [[epsilon].sub.r] (%), strain at break 100 5 [[sigma].sub.r] (MPa), stress at break 65-70 72 [lambda] (W [m.sup.-1] [K.sup.-1]), thermal conductivity 0.21 - d, density at 23[degrees]C 1.2 - TABLE 2. Mechanical and physical properties of the compatibilizers E-EA-MAH and E-MA-GMA. Melt Melting Vicat Index temp. temp. Base Type (dg [min.sup.-1]) ([degrees]C) ([degrees]C) E-MA-GMA AX8920 5 77 40 E-EA-MAH AX4700 6 63 40-104 Testing method ASTM D1238 D. S. C. ASTM D638 Maleic Glycidyl Ester Acrylate anhydride methacrylate Base (mol. %) (mol. %) (mol. %) (mol. %) E-MA-GMA 16 - - 1 E-EA-MAH 25 28 3 - Testing method F.T.I.R. F.T.I.R. F.T.I.R. [[sigma].sub.r] stress [[epsilon].sub.r] strain at break at break Base (MPa) (%) E-MA-GMA 8.6 800 E-EA-MAH 4 1100 Testing method ASTM D638 ISOR527 ASTM D 638 TABLE 3. Comparison of stress, strain, and modulus of rPC-20% CR w/w composites as a function of CR treatment. Untreated Untreated rPC-20% CR [[phi].sub.mean] CR [[phi].sub.mean] w/w CR = 140 [micro]m = 315 [micro]m [[sigma].sub.r] (MPa), stress 40 16 at break E (MPa), Young's 1500 950 modulus [[epsilon].sub.r] (%), strain 7 1.25 at break Washed Flamed rPC-20% CR [[phi].sub.mean] CR [[phi].sub.mean] w/w CR = 315 [micro]m = 315 [micro]m [[sigma].sub.r] (MPa), stress 36 34 at break E (MPa), Young's 1500 2000 modulus [[epsilon].sub.r] (%), strain 2.28 2.5 at break
|Printer friendly Cite/link Email Feedback|
|Author:||Zribi, K.; Elleuch, K.; Feller, J.F.; Bourmaud, A.; Elleuch, B.|
|Publication:||Polymer Engineering and Science|
|Article Type:||Technical report|
|Date:||Nov 1, 2007|
|Previous Article:||Photoinitiated crosslinking of Ethylene-vinyl acetate copolymers and characterization of related properties.|
|Next Article:||Dynamic viscoelastic properties of fluoroelastomer/clay nanocomposites.|