Mixed peroxide-sulfur rubber curing system.

The traditional curing system for rubber compounds consisting of sulfur and cure accelerator provides a good balance of mechanical properties and cost compared to other curing systems. However, the system has some disadvantages such as:

* Cure reversion at high curing temperatures;

* Relatively low stability of crosslink bonds between polymer chains (the polysulfidic nature of the bonds and relatively low bond energy - [congruent] 251 kJ/mol compared to 352 kJ/mol for a C-C bond) (ref. 1). This is one of the reasons for poor aging resistance and high compression set of cured rubber compounds.

* Environmental concerns related to nitrosamines (suspected carcinogens) formation associated with many of cure accelerators during rubber mixing and cure operations (ref. 2).

Peroxides are used in rubber applications where aging resistance, compression set or color stability are important factors. Whereas rearrangements can take place between polysulfide crosslinks, cyclic sulfides and free sulfur in a sulfur cured vulcanizate on aging, no such changes are possible in the C-C crosslinks of a peroxide vulcanizate. In the latter case, only chain scission can occur on aging.

The effect of these two curing systems on some rubber properties are summarized in table 1. In some cases, the effect is quite opposite. For example, during heat aging the sulfur curing system provides increased crosslink density whereas the peroxide system usually gives a decrease in the crosslink density. The curing behavior of compounds with the two curing systems is also different (figure 1). The peroxide cured compound shows lower scorch safety and cure rate but has no cure reversion. On the contrary, the sulfur cured compound has good scorch safety and high cure rate but shows a cure reversion.

Table 1 - features of rubber curing systems
                             Sulfur/    Peroxide
                         accelerator
Scorch safety:                  Good         Low
Cure rate:                      Fast        Slow
Cure reversion:                  Yes          No
Mechanical properties:          Good      Medium
Aging resistance:               Poor        Good
Fatigue life:                   Fair   Excellent
Compression set:                High         Low
Environment problems:            Yes          No

This work was initiated to develop a mixed curing system which would combine the advantages of each system and eliminate, if possible, the disadvantages.

The use of peroxide-sulfur mixed curing systems has been reported in the past (refs. 1, 3-32). There were indications that the tensile properties of carbon black filled EPM (ethylene-propylene copolymer) can be improved by the addition of a small mount of sulfur (refs. 3-6).

Some researchers used a mixed system for co-vulcanization of incompatible elastomers such as NBR (nitrile butadiene rubber) and EPM or EPDM (refs. 7 and 8) (ethylene-propylene-diene terpolymer); NR (natural rubber) and EPDM (refs. 9-13) for ozone resistance improvement: NR and cellulose short fibers for bonding them together (ref. 14). There are papers describing the use of a peroxide-sulfur system to improve the quality of some rubber articles : insulating compounds for electrical cables with reduced or no tracking (refs. 15 and 16), sponge rubbers with reduced density and compression (ref. 17) and vibration dampers with reduced stiffness (ref. 18). A mixed curing system can improve some rubber properties: adhesion to metal and nylon fibers (refs. 19 and 20); cut growth resistance and reduced heat build-up of a tread compound (refs. 7, 8, 21-24).

The mechanism of peroxide-sulfur cure has been discussed in several publications (refs. 1, 6, 25-32). It was suggested that sulfur can react with a polymer radical formed after the alkoxy radical abstracted a hydrogen atom from a polymer, creating persulfenyl radicals. Recombination and disproportionation of polymeric and persulfenyl radicals (ref. 8) lead to formation of polysulfidic crosslinking in addition to polymer-polymer C-C- crosslinking as a result of two polymer radicals interaction (chain termination) (refs. 1 and 6).

In spite of relatively large number of published articles on mixed peroxide-sulfur curing systems, we could not find papers discussing the effect on the cure systems on a variety of mechanical properties of the same compound including fatigue life, compression set, aging resistance and hysteresis.

In this article we have summarized the results obtained using a published tire sidewall formulation based on a blend of EPDM, BR (butadiene rubber) and NR (ref. 33). The compound was cured with a variety of mixed peroxide-sulfur curing systems and tested to evaluate a wide set of mechanical properties.

Experimental

Materials

A tire sidewall compound based on blend of EPDM, BR and NR was used in the experiments. Both the EPDM (Polysar 585) and the polybutadiene (Taktene 1203) are products of the Polysar Rubber Corp. The formulation, with exception of the carbon black used, was taken from W. von Hellens publication (ref. 33). The formulation is presented in table 2. The carbon black used was N330 from J.M.Huber Corp. rather than N660 used by von Hellens. The peroxide was [alpha],[alpha]'-bis(t-butyl peroxy) diisopropyl benzene supported at 40% concentration on Burgess KE clay (Vulcup 40KE peroxide). This is a product of Hercules Inc.

Table 2 - compound formulations(*)
                    Control   -1    -2     -3
EPDM 585              30
BR 1203               20
NR (SMR-5)            50
Stearic acid           2
ZnO                    3
Carbon black N330     50
Naphthenic oil(**)     20
Peroxide(***)                 1.5   2.0    2.5
Sulfur               1.75     1.3   1.0    0.5
Accelerator CBS      1.0      1.0   0.75   0.75
(*) The basic formulation was taken from ref. 33
The carbon black is different.
(**) Circosol 4240, Sun Refining Co.
(***) Vulcup 40KE, Hercules Inc.

Mixing

All compounds were mixed in a laboratory internal mixer - size BR, 77 rpm. The ingredients except curatives were charged into the mixer according to the following procedure:

First stage: 0' - all rubbers + 1/2 carbon black + stearic acid;

               0.5' - 1/2 carbon black + oil + zinc oxide;
               4.0' - sweeping;
               5.0' - dumping.

Second stage: Curatives were milled in for two minutes on an open lab mill heated to 40 [degrees] C. Final compound was passed six times through 0.5 mm nip setting on the mill.

Curing

All samples were pressed cured at 182 [degrees] C. The cure time was determined using an oscillating disk rheometer according to the ASTM D2084 procedure. The measurements were made at 182 [degrees] C using a [plus or minus] 3 [degrees] arc and 12' motor. Approximately 1.5 min. time was added to the [t.sub.95] determined by the rheometer tests for press mold heating.

Testing

Physical properties were determined according to the ASTM D412 procedure. Samples were cut from cured 150x150 mm slabs having thickness [congruent] 2 mm with a standard Die C. The tear resistance was determined according to ASTM D624, Die C procedure. The Shore A durometer was used to determine the hardness (ASTM D2240). Heat aging resistance was determined according to ASTM D573 at 100 [degrees] C. The compression set was determined according to ASTM D35, Method B, under constant deflection in air using direct molded disk samples.

Fatigue life was determined by using a fatigue-to-failure tester, Cam #10 (extension 89%) and following the procedure outlined in its operation and service manual.

The hysteresis of rubber compounds was evaluated by measuring the ratio of lost strain energy to the input (total) energy as it is described in some national standards (ref. 34). The energies were measured after the first and fifth cycle of extension to 100% elongation. The tests were performed on standard samples and rate of grip separation (Die C, ASTM D412) by using the Monsanto T100 tester. The parameter characterizing the hysteresis properties was called "relative hysteresis."

A statistically designed experiment was conducted to optimize the amount of sulfur, accelerator CBS and peroxide in order to obtain curative packages giving rubber compounds with approximately the same crosslink density and different peroxide to sulfur/accelerator ratios. A 12-run Box-Behnken design with five replicated center points was used to generate the data. This design allows an estimation of main factors effects, two-factor interactions and curvature effects of the three variables - peroxide, sulfur and cure accelerator.

Results and discussion

All investigations in this work were done using the formulation shown in table 2. To be accepted in the tire industry, a tire sidewall compound should meet the requirements at least for the following properties:

* Processability;

* Curing behavior;

* Physical properties;

* Aging resistance;

* Stability of crosslink bonds;

* Hysteresis;

* Fatigue life;

* Adhesion to carcass and tread;

* Ozone resistance.

All the properties mentioned above except the adhesion to rubber carcass and tread and ozone resistance were evaluated in this work.

In the patent issued to Uniroyal Chemical (ref. 24) it was shown that the mixed peroxide-sulfur/accelerator curing system exhibits an improved adhesion of a tire sidewall to adjacent tire parts. In the published papers (ref. 33 and 37) it was shown that EPDM in a blend with NR and butadiene elastomer provides an adequate ozone protection if the amount of EPDM is more than 30 wt.% of all elastomers. This is one of the reasons for EPDM use in a tire sidewall.

Based on model equations and contour plots from the design study, we have chosen for the investigation three compounds with different ratios of peroxide to sulfur/accelerator curing systems which have crosslink densities (ODR [Delta] torque) close to the control compound cured with a traditional sulfur/accelerator system.

Processability

Normally, processability is evaluated by two major parameters: Mooney viscosity and scorch safety. Because the basic formulation was the same and we assumed that the curatives will not affect Mooney viscosity, we did not run the tests. The scorchiness was evaluated by the parameter [t.sub.s]2 from the ODR curves.

Curing behavior

Table 3 and figure 2 show the results obtained. As expected, the scorch safety ([t.sub.s]2) and cure rate of the compounds cured with a mixed curing system (M compounds) are lower compared with the sulfur/accelerator cured compound (S-compound or control compound). The difference in the two properties and also the reduction in cure reversion are proportional to the amount of peroxide used. Significant reduction or elimination of cure reversion can allow an increase in the curing temperatures which results in increased productivity in case the scorch safety is adequate.

Table 3 - curing behavior of compounds cured
with a mixed system
                            Control    -1     -2     -3
Curing ingredients:
Sulfur                         1.75   1.3    1.0    0.5
Accelerator CBS                 1.0   1.0   0.75   0.75
Peroxide                              1.5    2.0    2.5
ODR, 182 [degrees] C(12')
Max. torque, dN.m              52.9  55.0   50.1   50.5
Min. torque, dN.m               8.0   8.0    7.7    7.8
Delta torque, dN.m             44.9  47.0   42.4   42.7
[t.sub.S] 2, min.              1.45   1.3   1.17    1.1
[t.sub.95], min.                2.9   3.7    4.7    6.0
Reversion, dN.m(-)              7.0   1.0    0.5    0.2

Physical properties

From the data in table 4 we can see that in spite of some differences in the crosslink density (ODR [Delta] torque, modulus at 100% elongation) the M-compounds have higher tensile strength and equal or better tear resistance. The elongation at break for all compounds is essentially the same. It should be noted that although the moduli at 100% elongation and hardness of the M-compounds are equal or lower than the ones of the S-compound, the moduli at 200% and 300% elongation of all M-compounds are higher.

[Tabular Data Omitted]

Aging resistance

A big difference in aging resistance between the M-compounds and the S-compound as found, especially in aging for 168 hrs. at 100 [degrees] C (table 5). The difference is more pronounced in the elongation at break. The parameter is only 90% for the S-compound and 148-300% for the M-compounds. The more visible way to evaluate the change in properties after aging is the use of so-called "Aging coefficients" which are the ratio of a property value after aging to the original value. From the data in table 6 we can see that the moduli change occurs to a lesser extent with an increase in the peroxide to sulfur/accelerator ratio. The elongation at break retention after aging is increasing with the amount of peroxide used.

[Tabular Data Omitted]

Crosslink stability

Mechanical properties of rubber compounds significantly depend on the crosslink density and the nature of the crosslink bonds. Mechanical stresses and high temperatures acting for a long period of time can cause big changes in the crosslink network, usually leading to reduced service life of rubber articles. Cunneen and Russell (ref. 35) showed, for example, that during tire service significant changes occurred in the chemical structure of the tread vulcanizate. These included a marked reduction in the concentration of polysulfide bonds and an increase in the amount of monosulfide crosslinks. This was due to thermal anaerobic reactions. As the result of these changes, some of the physical properties of the tread compound deteriorated. The change in properties can lead to reduction on service life of a tire (or another rubber article). Increasing stability of the crosslink network will improve the quality of rubber articles and their service life.

The effect of mechanical stresses and heat on the stability of network structures can be evaluated by measuring the compression set. Figure 3 shows the result obtained. It is obvious that the M compounds have a significant advantage in crosslink stability over the control compound. The more peroxide used, the greater the advantage.

Hysteresis

For a tire sidewall the heat developed due to hysteresis causes deterioration of sidewall adhesion to the carcass. The tear resistance and fatigue life are reduced also. So, the hysteresis should be as low a possible. The data in figure 4 show that the M-compounds have lower hysteresis compared with the S-compound. This is in agreement with the data presented in the Uniroyal Chemical patent (ref. 24).

Fatigue life

Fatigue life of a tire sidewall compound is probably the most important property. The sidewall must withstand an enormous amount of cycles during the tire service life. In some cases fatigue life determines the tire durability. The sidewall is exposed to air (oxygen), ozone, heat, moisture and sunlight. It must be adequately protected. For this reason, the use of EPDM in a tire sidewall was an objective of some patents (refs. 24 and 36) and publications (refs. 33 and 37). The data in table 7 show clearly the superiority of M-compounds over S-compound in fatigue life both of unaged and, especially, aged samples.

[Tabular Data Omitted]

Environmental problems

The mixed curing systems use less nitrogen containing cure accelerators compared with a sulfur curing system. Therefore, the possibility of nitrosamine generation associated with these accelerators is reduced.

Conclusions

Compared with a traditional sulfur curing system the mixed peroxide-sulfur/accelerator system can provide:

* Elimination (significant or reduction) of cure reversion at high temperatures;

* Equal or better physical properties;

* Significantly improved aging resistance, especially retention of elongation at break;

* Significantly reduced compression set;

* Lower hysteresis;

* Significantly increased fatigue life, especially after aging;

* Safer environment.

References

[1.] S.K. Chakraborty, A.K. Bhowmick and S.K. De, J. Macro. Sci.-Revs. Macro. Chem. C21 (2) 313-332 (1981-82). [2.] T.L. Jablonowski, Rubber World, 206 (5) 18-22) 1992. [3.] A.E. Robinson, J.V. Marra and L.O. Amberg, Preprints, Symposium on Synthetic Elastomers from Petroleum Hydrocarbons, 6, 4A, A-117 (1961) [4.] L.O. Amberg, Fourth Rubber Technology Conference, London, 1962, p. 527. [5.] L.O. Amberg, Polymer Letters, 2, 59-62 (1964). [6.] L.D. Loan, J. Pol. Sci., Part A. 2, 3053-3066 (1964). [7.] D.E. Wingrove (to Uniroyal Chemical Co., Inc.), EP 229, 489 (July 22, 1987); Chem. Abstr. 107 (12) 98003d. [8.] D.E. Wingrove (to Uniroyal Chemical Co., Inc.) U.S. 4,743,656 (May 10, 1988). [9.] Shinoda (to NOK Corp.), J.P. 61/293, 240 (December 24, 1986); Chem. Abstr. 107 (8) 60643r. [10.] S.D. Tobing, Rubber World. 197 (5) 33-37 (1988). [11.] S.D. Tobing (to Servus Rubber Co., Inc.) U.S. 4,931,508 (June 5, 1990). [12.] Yasumoto (to Sanshin Chemical Industry Co., Ltd.) J.P. 62/169,826 (July 27, 1987); Chem. Abstr. 107 (20) 177821x. [13.] Uniroyal Inc., J.P.61/14239 (January 22, 1986); Chem. Abstr. 105 (12) 98984h. [14.] P. Flink, B. Westerlin, M. Rigdahl and B, Stenberg, J. Appl. Pol. Sci. 35 (8) 2155-64 (1988). [15.] Showa Electric Wire and Cable Co., Ltd., J.P. 56/109, 402 (August 29, 1981); Chem. Abstr. 94 (26) 221113c. [16.] E.P. Solov'ev, A.V. Kuznetsov, R.V. Kuznetsova, V.O. Reikhsfel'd and V.I. Klochkov, Izv. Vyssh. Uchebn. Zaved., Energ., (1), 35-37 (1984). [17.] M. Aoshima et al (to Sumitomo Chemical Co., Ltd) J.P. 62/277, 442 (December 2, 1987); Chem. Abstr. 108 (26) 222850x. [18.] T. Kawada et al (to Japan Synthetic Rubber Co., Ltd.) J.P. 63/243, 153 (October 11, 1988); Chem. Abstr. 110 (8) 59305w. [19]. Y. Itokazu et al (to Bando Chemical Industries, Ltd.) J.P. 63/221144 (September 14, 1988); Chem. Abstr. 110 (8) 59323a. [20.] O. Ozawa et al (to Yokohama Rubber Co. Ltd.) J.P. 63/159454 (July 2, 1988); Chem. Abstr. 109 (22) 191963t. [21.] J.M. Mitchell et al (to Uniroyal Chemical Co., Inc.) U.S. 4,814,384 (March 21, 1989). [22.] J.M. Mitchell et al (to Uniroyal Chemical Co., Inc.) E.P. 342, 288 (November 23, 1989); J.P. 1,314,601 (December 19, 1989. [23.] D.E. Wingrove (to Uniroyal Chemical Co., Inc.) U.S. 4,767,809 (August 30, 1988). [24.] J.M. Mitchell (to Uniroyal Chemical Co, Inc.) U.S. 4,973,627 (November 27, 1990). [25.] P. Viohl, J.R. Little and C.H. Stiteler, Rubber Age 94, 594-597 (1964). [26.] C.K. Das and S. Banerjee, J. Polym. Sci., Polymer Chem. Ed. 16,2971-2983 (1978). [27.] S.P. Manik and S. Banerjee, Rubber Chem. Technol. 43 (6) 1311-26 (1970). [28.] S.P. Manik and S. Banerjee, J. Appl. Polym. Sci. 15 (6) 1341-55 (1971). [29.] P.K. Bandyopadhyay and S. Banerjee, Kautsch. Gummi, Kunst. 32(12) 961-7 (1979); Chem. Abstr. 92 (10) 77765k. [30.] A.K. Ghosh, C.K. Das and S. Banerjee, J. Polym. Sci., Polym. Chem. Ed. 15(11) 2773-84 (1977). [31.] C.K. Das, A.K. Ghosh and S. Banerjee, J. Polym. Sci., Polym. Chem. Ed. 15 (9) 2255-68 (1977). [32.] C.K. Das and S. Banerjee, Rubber Chem. Technol. 47 (2) 266-81 (1974). [33.] W. von Hellens, Paper #40, presented at the 136th ACS Rubber Division Meeting, Detroit, Michigan, October, 1989. [34.] Soviet Standard. GOST 252-75. [35.] J.I. Cunneen and R.M. Russell, Rubber Chem. Technol.43 (5) 1215 (1970). [36.] W. von Hellens et al. (to Polysar Ltd.) U.S. 4,645,793 (February 24, 1987). [37.] W. von Hellens et al., Rubber and Plastics News, 1990 Technical Yearbook, 88-92.