Redefining thermosets: a method of molding vulcanized rubber into new parts.

For many years, researchers have been searching for a solution to the problem of recycling thermosetting materials (crosslinked polyurethanes, natural and synthetic rubbers, etc). Most polymer science textbooks classify polymers into two categories: thermoplastic and thermoset. Thermosetting materials are defined as thermoplastic or liquid precursor polymers that react to form crosslinked networks upon vulcanization. Charles Goodyear and Nathaniel Haywood first discussed sulfur vulcanization in patents between 1842-1844. In his patent Goodyear writes that, "No degree of beat without blaze can melt it (rubber) ..." The patent also states that rubber is resistant to the most powerful chemicals of the time period (refs. 1-3). With sulfur vulcanization, Goodyear in vented one of the most important materials of the past several centuries. Unfortunately, he also created a thermosetting material that has become a recycling nightmare. Ironically, in 1853 he realized this problem and patented a process to shred the scrap-vulcanized rubber (ref. 4). The shredded/powdered vulcanized rubber was blended into uncrosslinkcd rubber and then vulcanized. This technique is still used today and is known as regrind blending. In 1855, Charles Morey patented a process that he described as, "Forming or molding scrapings, filings, dust, powder or sheets of hard vulcanized India-rubber into a compact solid mass by means of a high degree of heat and pressure ..." (ref. 5). This concept was revisited in 1981 when Accetta el al. (refs. 6 and 7) described molding ground rubber tires at high temperatures and pressures while incorporating additional vulcanizing agents, such as sulfur. Law et al. (ref. 8) discussed a proprietary method of molding polyurethane powders into films and studied the effect of particle size on mechanical properties. Arastoopour et al. (ref. 9) have received a patent for "processing recycled rubber." This document presents little mechanical data and the majority of the patent is focused on rubber grinding. Adhikari et al. (ref. 10) offers an excellent review of the past and present techniques for rubber recycling and lists numerous references for further study. In particular, there has been a tremendous amount of work on chemical methods of depolymerization/ devulcanization of rubbers to yield value-added chemicals from scrap rubber (ref. 10).

Americans dispose of approximately 275 million tires per year, and this is only a fraction of the total amount of thermosets discarded annually (ref. 11). Many of these tires are subsequently composted in landfills where they become major health risks. The potential dangers include the pollution of soil and water, fire hazards and the tires acting as breeding grounds for mosquitoes and rodents, and the diseases associated with them (ref. 12). Presently, many landfills will not accept scrap tires for these reasons. There are currently two major uses of scrap tires. The first use is to burn the scrap tires for fuel. Tires have a fuel content equivalent to coal, however, in some cases this procedure can be more polluting than burning fossil fuels. The second reuse involves grinding the rubber into a powder and incorporating this powder as a low volume filler into materials such as asphalt, unvulcanized rubber and concrete. Unfortunately, virtually no scientific research has been conducted in the area of producing new rubbers directly from post-vulcanized (scrap) rubbers without first depolymerizing the rubber, which is a difficult and environmentally polluting process. The abovementioned techniques are simply methods of reducing the burden rubber waste has on our nation's landfills (refs. 10-16).

High-pressure, high-temperature sintering (HPHTS)

Until recently, there has been no means of taking 100% waste rubber and making new rubber products. This is the main advantage of high-pressure, high-temperature sintering (HPHTS). The process takes rubber powder, ground from waste rubber, and converts it into new products. The technique works for practically every common rubber, and in many cases yields parts with good mechanical properties. Figure 1 depicts the "reknitting" of the particles by HPHTS. The individual particles recrosslink, fusing the interfaces to form a new network. Afterward, the particle interfaces are indistinguishable from the bulk material. Pressure is needed for intimate contact of the particles to occur. Temperature adds the necessary energy to break the existing crosslinks, and allows them to reform throughout the material. It is believed that the mechanism of particle adhesion is the chemical exchange reactions (formation of free radicals which then recombine) that occur in rubbers at elevated temperatures. These chemical interchange reactions have been known for many years and date back to work done by A.V. Tobolsky in the 1950s (refs. 17-20). Chemical stress relaxation is defined as the mechanical relaxation of stresses caused by the exchanging of chemical bonds in a network. Both intermittent and continuous chemical stress relaxation yield information about the crosslink network, which undergoes changes at elevated temperatures. These techniques can be used to measure the destruction of the original network, as well as the formation of the new network at elevated temperatures. Interchange reactions hold the key to understanding and optimizing rubber sintering.

[FIGURE 1 OMITTED]

Of particular interest is the modulus of the material. The modulus yields information about the crosslink density of the overall rubber network. Figure 2, from Tobolsky and MacKnight, shows the relative modulus versus time curves at 140[degrees]C for natural rubber (NR), polysulfide rubber (also known as Thiokol rubber [TR]), and styrene-butadiene rubber (SBR). The relative modulus for polysulfide rubber is constant with a value of one. This correlates to the formation of one bond lot every bond that breaks (i.e., constant crosslink density). On the other hand, the relative modulus for NR decreases and the modulus for SBR increases. Therefore, NR is reverting to an uncrosslinked material (i.e.. losing crosslinks faster than they are reforming), while SBR is forming new crosslinks faster than they are breaking. This is extremely important, as work by Gent in figure 3 illustrates that there is a parabolic-like correlation of strength and crosslink density (ref. 21). Most rubbers are formulated to yield the maximum obtainable strength. This strength will decrease by either increasing or decreasing the total crosslink density, correlating to over-cure and reversion, respectively. The phenomena of reversion and over-cure have been studied for many years. There are many publications discussing materials which slow/stop the reversion/interchange reactions, thus creating more thermally stable rubbers (refs. 22-25). Additionally, M. Bellander, et al., have shown that polybutadiene rubber is capable of self-vulcanization at high temperatures while polyisoprene is not (ref. 26). These data further supports the idea that rubber containing butadiene will over-cure at high temperatures, while lubbers containing isoprene will revert.

[FIGURES 2-3 OMITTED]

Understanding how different rubbers react at high temperatures is a very important point when considering rubber recycling via HPHTS. Table 1 shows the sintered mechanical properties obtained for three types of vulcanized rubbers: NR, SBR and TR. The vulcanized properties were optimized to yield the maximum strength, and were cured at 90% of the optimum time ([t.sub.90]). The mechanical properties of both sintered NR and SBR are less than those of the starting materials, while the sintered TR can achieve 100% recovery of the virgin properties. This is due to the trends shown in figures 2 and 3 from Tobolsky et al., and Gent, respectively. As previously stated, figure 2 yields information about the crosslink density after exposure to elevated temperatures where exchange reactions occur. If the relative modulus is greater than unity, the crosslink density is increasing, and if the relative modulus is less than unity, the crosslink density is decreasing. Since NR reverts (reducing the crosslink density) and SBR over cures (increasing the crosslink density) the final strength of these materials should decrease due to the change in network structure. This has been observed experimentally as shown in figures 4a, b and c. On the other hand, TR has a constant modulus and constant crosslink density. TR should, therefore, recover 100% of the properties. Experimentally, 94% of the original strength has been obtained.

[FIGURE 4 OMITTED]

Experimental and results

Figures 4a, b and c show the effect of the molding temperature on the mechanical properties of NR, SBR and TR, respectively. The strength and modulus are plotted on the left hand axis, while the elongation at break is shown on the right hand axis. The solid data points represent the heat-treated sheets and the hollow points represent the sintered powder that was derived from these sheets prior to heat-treatment. The sintered parts and heat-treated sheets were molded at 1,250 psi (~8 MPa) for one hour. The symbols on the figures correlate as follows: Strength is represented by squares, elongation at break by circles and the modulus by diamonds. Figures 4a and b show a decrease in the strength of the heat-treated sheets as the molding temperature increases. The decrease in strength is due to the change in crosslink density that lowers the ultimate strength, as shown in figure 3. The strength of the sintered NR, SBR and TR increases with increasing temperature until the sintered curves intersect the heat treated sheet curves. At these points, the sintered rubber powder and heat treated rubber sheets are mechanically identical. Since the sintered properties are equal to the heat treated sheets at these intersections, the limiting variable in obtaining 100% recovery of the original mechanical properties is the changing of the crosslinked network at elevated temperatures. Therefore, the percent of the strength that is recoverable is related to the crosslink density. For TR rubber, it is possible to obtain 100% of the starting properties, while NR and SBR rubbers recover ~60% of the starting properties. The 100% modulus, a commonly used term in discussing the properties of rubbers, is proportional to the crosslink density. Thus, the crosslink density of NR is monotonically decreasing above 160[degrees]C, while the crosslink density of SBR is monotonically increasing above 230[degrees]C. TR has a constant modulus until approximately 160[degrees]C (sulfur bonds begin to break at 80[degrees]C) where reversion begins. By plotting these data as strength versus crosslink density (similar to figure 3), it is possible to further show this trend. From these figures it is evident that above the critical temperature the heat treated sheets and the sintered powders are identical.

Figures 5a, b and c show the strength versus crosslink density for NR, SBR and TR, respectively. A measure of the crosslink density was obtained from swelling in toluene, a commonly used method to evaluate this parameter in rubbers. The value Q represents the swelling ratio defined as the swollen mass over the initial mass. In the strength versus crosslink density of a conventionally vulcanized rubber, the strength behaves similar to the trend shown in figure 3. For the three rubbers, the ground powder has the same crosslink density as the starting sheets, but there is no adhesion between particles. This correlates to the zero strength data point in figures 5 a, b and c. In figure 5a, the crosslink density of NR is decreasing as the strength of the sintered rubber first increases and then plateaus. Finally, the strength decreases until the line intersects the dotted curve of the heat-treated sheets "master curve" (ref. 27). The exact curvature of the line is unknown, but the trend of decreasing strength with increasing and/or decreasing crosslink density is definite and is based on our previously unpublished work and Gent's work (ref. 21). Additionally, figure 5a shows the effect of 1,3 biscitraconimidomethyl benzene (Perkalink-900), an anti-reversion agent, on the crosslink density of NR. Thesr data summarize our submitted work on the effect of additives on the mechanical properties of sintered rubbers (ref. 28). With the addition of the anti-reversion agent. the crosslink density and strength are increased as the anti-reversion agent reforms bonds broken during reversion. In figure 5b, the crosslink density of sintered SBR is constant until 230[degrees]C. This temperature is approximately the autovulcanization temperature of butadiene (ref. 26). Above this temperature, the strength continues to increase as the crosslink density increases until intersecting the "master curve." The "master curve" was obtained from heat-treating rubber sheets, and then measuring both the strength and crosslink density. The strength of TR sintered rubber increases at constant crosslink density, recovering 100% of the strength. Above 160[degrees]C, the crosslink density decreases, and therefore there is a decrease in the strength. This decrease is identical to the decrease in the strength of the heat-treated sheets as shown in figure 5c. Nonetheless, there is a small processing window where it is possible to recover 100% of the starting properties. Therefore, it is believed that by creating a rubber that has a constant crosslink density independent of the exchange chemistry that occurs at high temperature, it should be possible to achieve 100% recovery in the properties. Such a rubber might be produced by co-polymerizing isoprene (decrease in modulus) and butadiene (increase in modulus).

[FIGURE 5 OMITTED]

Finally, HPHTS is capable of creating some unique rubber blends, which would be conventionally impossible to produce due to the incompatible nature of certain combinations of different types of rubber. HPHTS allows for the blending of materials that would normally phase separate in their uncrosslinked liquid state. In this study, the two rubbers investigated are natural rubber and Viton, a fluoroelastomer. These two rubbers are two of the most difficult rubbers to blend due to the inherent differences in their chemical backbone. Nonetheless, the two vulcanized rubber powders are capable of being sintered together without any loss in the mechanical integrity. Figure 6 shows the mechanical properties of a blend of the sintered rubber powder. The size of the resulting rubber phases is controlled by the starting particle size and the blend ratio.

[FIGURE 6 OMITTED]

Conclusions

High-pressure, high-temperature sintering (HPHTS) is a viable technique to recycle the large amounts of unwanted scrap rubber generated daily. This unique process is very simple and is easily scaled up as it uses standard rubber processing equipment (compression molding). We have demonstrated that the most commonly used rubbers are capable of being sintered back together from their powdered state and that this process can turn scrap tires into a new, valuable feedstock. Additionally, the traditional definition of thermosetting materials may have to be rewritten. In our opinion, the underlying chemistry for this process should open new avenues for rubber recycling, as well as the recycling of other thermosetting materials.

Table--1 mechanical properties of some common rubbers
(starting properties in parenthesis)

Type of rubber             Sintering conditions      Strength
                                                       (MPa)

SBR                     240[degrees]C 1 hour 8 MPa    15 (24)
NR with Perkalink 900   180[degrees]C 1 hour 8 MPa     9 (15)
NR                      180[degrees]C 1 hour 8 MPa     7 (15)
Polysulfide             140[degrees]C 1 hour 8 MPa   7.5 (7.9)

Type of rubber          Elongation at   100% modu-   Compression
                          break(%)       lus(MPa)

SBR                      303 (650)      2.3 (1.6)       Same
NR with Perkalink 900    221 (250)      2.5 (1.6)        NA
NR                       280 (250)      1.3 (1.6)        NA
Polysulfide              247 (236)      3.1 (3.4)        NA

References

(1.) C. Goodyear U.S. Patent 3,633 (1844).

(2.) K. Bradford and D.D. Pierce, "Trials of an inventor, life and discoveries of Charles Goodyear," Phillips & Hunt Publishers, New York, 1866.

(3.) N. Hayward, U.S. Patent 1,090 (1839).

(4.) C. Goodyear, Br Patent 2,933 (1853).

(5.) C Morey, U.S. Patent 12,212 (1855).

(6.) A. Accetta and J.M. Vergnaud, Rubber Chemistry and Technology, 54, 302 (1981).

(7.) A. Accetta and J.M. Vergnaud, Rubber Chemistry and Technology, 55, 961 (1982).

(8.) W.K. Law, T. Patel, K. Swisher and F. Shutov, Polymer Recycling 3, 269 (1998).

(9.) H. Arastoopour, D. Schoke, B. Bernstein and E. Bilgili, U.S. Patent 5,904,885, 1999.

(10.) A. Adhikari, D. De and S Maiti, Prog. Polym. Sci., 25, 909 (2000).

(11.) J. Jang, T. Yoo, J. Oh and I. Iwasaki, Resources, Conservation, and Recycling, 22, 1 (1988).

(12.) R.G. Stacer, Rubber Chem. Technol., 73, 551 (2000).

(13.) H. Morawetz, Rubber Chem. Technol. 73, 405 (2000).

(14.) J. Paul, Encyclopedia of Polymer Science and Engineering, Rubber Reclaiming, 14, 787 (1988).

(15.) J.A. Beckman, G. Crane, E.L. Kay and J.R. Laman, Rubber Chem. Technol., 51, 597, (1973).

(16.) W.H. Klingensmith and K.C. Baranwal, in "Handbook of Elastomers," Marcel Dekker, Inc., New York, 2001, chapter 34.

(17.) W.J. MacKnight and A.V. Tobolsky, "Polymeric sulfur and related polymers," Interscience Publishers, New York, 1965.

(18.) J.J. Aklonis and W.J. MacKnight, "Introduction to polymer viscoelasticity," Interscience Publication, New York, 1983.

(19.) A. V. Tobolsky, "Properties and structures of polymers," John Wiley and Sons, New York, 1960.

(20.) S. Tamura, K. Murakami and H. Kuwazoe, J. Appl. Polym. Sci., 28, 3,467 (1983).

(21.) F.R. Eirich, "Science and technology of rubber," Academic Press, New York, 1978.

(22.) P.J. Nieuwenhuizen, J.G. Haasnoot and J. Reeduk, Rubber Chem. Technol. 72, 15 (1999).

(23.) R.N. Datta, Rubber Chem. Technol. 72, 15 (1999).

(24.) R.N. Datta and J.C. Wagenmakers, J. Polym. Mater. 15, 379 (1998).

(25.) A.H.M. Schotman, et al., Rubber Chemistry Technology. 69, 727 (1996).

(26.) M. Bellander, B. Stenberg and S. Persson, Polymer Engineering and Science, 38, 1,254 (1998).

(27.) J. Morin, D.E. Williams and R.J. Farris, Rubber Chemistry and Technology in review.

(28.) A.R. Tripathy, J.E. Morin, D.E. Williams, S.J Eyles and R.J. Farris, Macromolecules in submission.