A chemical for reversion resistant compounding.

Reversion of sulfur based crosslinks continues to be a problem for compounders. To date, no ideal solution has been found to address reversion. With each new solution, has come a compromise; better heat and reversion resistance is generated at the expense of productivity, processing safety and/or fatigue and tear properties. To develop a solution, it is best to start with the definition of reversion.

Reversion is the thermal degradation of polysulfidic crosslinks leading to a reduction of crosslink density and an introduction of main chain modifications. In practice, reversion leads to a decline in compound physical properties. Modulus, resilience and other compound properties deteriorate as a consequence of reversion.

Reversion occurs when compounds are overcured or when vulcanizates are exposed to anaerobic againg. The reversion phenomenon is generally associated with the vulcanization of NR compounds at high temperatures or when compounds are cured for a prolonged time at more modest temperatures (ref. 1). While the practical significance of reversion lies in the effect on vulcanizate properties, the phenomenon can be most conveniently monitored by rheometer measurements.

A typical NR compound, rheograph showing reversion is given in figure 1 (ref. 2). The simulated parameter trace illustrates the three main regions of cure (labeled 1-3). The first region is the scorch delay period, in which most of the accelerator chemistry takes place. The second period is where the initial network structures are formed and the accelerator intermediates are consumed as crosslinks are developed. Finally, in the third stage, overcure and reversion occur. During this stage the network matures and reversion of polysulfidic crosslinks causes physical properties to deteriorate.

[ILLUSTRATION OMITTED]

Reversion during the service life of an article has also had to be addressed by compounders. Many compounds (OTR, truck, bus, racing, aircraft and high performance tire compounds) generate enough heat during service to cause degradation of the crosslink network. This process is self-perpetuating, since it lowers the modulus, which in turn accelerates the heat generation. The result of this process is premature failure and/or shortened service life. In the lab, flexometers can be used to detect irreversible changes to the compound caused by repeated cycling (ref. 3).

Since the loss of designed properties is undesirable, it is not surprising that various compounding methods have been developed to offset or reduce the effect of reversion. The simplest and often most used way to address reversion is to cure at lower temperatures. As figure 2 shows, curing a NR compound at 130[degrees]C leads to no reversion, while curing at higher temperatures leads to reversion. The problem with this method is that productivity suffers due to longer cure times. A compound cured at 170[degrees]C will reach optimum cure in 1/16 the time it takes for the same compound cured at 130[degrees]C. Additionally, this method does not address reversion during service, since the crosslinks generated polysulfidic and lack heat stability.

[ILLUSTRATION OMITTED]

To address reversion without reducing productivity, compounders have developed the use of efficient or semi-efficient vulcanization systems. These systems use high accelerator/sulfur ratios and/or sulfur donors to reduce the number of polysulfidic crosslinks generated. Reversion resistance is improved because the crosslink network that is developed is based on more heat stable di- and mono-sulfidic crosslinks. This improvement in reversion resistance is achieved, however, at the expense of scorch safety, flex fatigue and strength related properties. Additionally, compounds using efficient cure systems bond poorly to metal and fabric due to faster cure rates and lower total sulfur levels. These limitations have restricted the use of low sulfur and sulfur donor cure systems in compounds used in dynamic applications.

Another approach to prevent reversion is peroxide curing. Peroxide cure systems develop carbon-carbon crosslinks. These crosslinks are very heat-stable because they have a high dissociation energy (table 1). Due to the polymer chains being bonded through a carbon-carbon crosslink, peroxide systems give the same problem as efficient curing systems. Poor mechanical properties and little or no metal and fabric adhesion mean that these systems will see little use in dynamic applications.

Table 1 - bond energies of different crosslink types

Type of linkages                  Bond energies, kj [mole.sup.-1]
in rubber network
- c - [S.sub.x] - c - (x > 2)              < 268
- c - [S.sub.2] - c -                       268
- c - s - c                                 285
- c - c -                                   352

In search of an additive to protect against reversion, equilibrium cure systems (ref. 4) and specific post vulcanization additives (refs. 5 and 6) have been developed. Further, an investigation into the effect of zinc soaps of unreported composition has been reported (ref. 7). Recently, TBSI-accelerated vulcanization and vulcanization with (1,6)bis(N,N'-dibenzyl thiocarbamyl disulfide)hexane was reported (refs. 8 and 9). Although these systems partially serve the purpose, they are far from an optimum solution.

Lacking an ideal vulcanization system, semi-efficient cure systems have provided a compromise and their use has become widespread. The semi-efficient system, although exhibiting less reversion, remains inferior in dynamic performance compared with conventionally cured systems. Additionally, semi-efficient systems are of little use where metal and fabric adhesion is the target. The challenge remains to improve the reversion while maintaining all other performance characteristics.

Ideally, a system to address reversion would not affect compound properties except to eliminate reversion. Application of the new system would not alter scorch time, cure rate or optimum properties. The only effect on the compound would be a maintenance of properties under conditions that cause reversion.

Perkalink 900 (Pk900) is a new chemical that addresses reversion without changing compound properties. Scorch time, cure rate and optimum properties are not altered by the addition of Pk900. This is accomplished because Pk900 is essentially inactive until polysulfidic crosslinks begin to revert. Compound properties are maintained after reversion begins, but are unchanged when reversion is not present.

Experimental

Pk900 (figure 3) was evaluated in a reversion sensitive, NR compound. Compound recipes are shown in table 2.

[MATHEMATICAL EXPRESSION OMITTED]

Table 2 - natural rubber model compounds

Ingredients                             01         02
NR SMR CV                              100      100
Carbon black N-330                      50       50
Ar. oil ingralen 150                     3        3
Zinc oxide                               5        5
Stearic acid                             2        2
Perkacit CBS                             0.6      0.6
Sulfur                                   2.3      2.3
Perkalink 900                            -        1.0
Cure time
[t.sub.90], min. (150[degrees]C)        13       13
[t.sub.90], min. (170[degrees]C)         3.2      3.2

The experimental compounds were mixed in two stages. In the first phase rubber, carbon black, oil and all other ingredients except CBS, sulfur and Pk900 were combined. The compound was mixed for five minutes in an internal mixer; the dump temperature was 130-135[degrees]C. The CBS, sulfur and Pk900 were added to the masterbatch on a two-roll mill.

Testing

Cure properties for the compounds were determined using a Monsanto MDR 2000E at 150[degrees]C and 170[degrees]C. Vulcanizates were cured under four conditions: 150[degrees]C x t90, 150[degrees]C x overcure, 170[degrees]C x t90 and 170[degrees]C x overcure.

Physical testing procedures used are as follows:

* Stress strain data to IS0 37/2

* Tear strength (crescent, 1 mm cut) to IS0 34

* Heat build up, Goodrich Flexometer to IS0 4666/3 Load: 11 kg, stroke: 4.45 mm, freq.: 30 Hz

* Abrasion, DIN to IS0 4649

* Compression set, IS0 R815, 3d/100[degrees]C

* Fatigue to failure (Monsanto), ASTM 4482/85, Cam: 24

* Dynamic mechanical analysis, Rheometric dynamic analyzer (700): Frequency 15 Hz, 60[degrees]C, dynamic strain 0.75%

* Aging experiments - aging conditions: 100 [+ or -] 1[degrees]C for 24-72h.

Structural characterization

The crosslink density and the distribution of poly-, di- and mono-sulfidic and non-sulfidic crosslinks were determined. A natural rubber gum recipe (NR 100; stearic acid 2; zinc oxide 5; Perkacit CBS 0.6; and sulfur 2.3 phr) was used for this determination. The compound was mixed on a two-roll mill and then vulcanized under two conditions: 150[degrees]C x [t.sub.90] (optimum) and 170[degrees]C x 30 min. (overcure).

Before vulcanization, the number average molecular weight of the different stocks was determined from intrinsic viscosity measurements (ref. 10). An average value of 1.4 x [10.sup.5] was determined in this study. The crosslink density in gmole/g-rubber hydrocarbon was estimated from the elastic constant using Mullins relationship (ref. 11). The value of the elastic constant from the vulcanizates was determined from the Mooney-Rivlin expression (ref. 12) following the procedure given by Saville and Watson (ref. 13).

The proportions of mono-, di- and polysulfidic crosslinks in the vulcanizate were determined by thiol-amine chemical probe (ref. 14). After poly- and di-sulfidic linkages were degraded, the samples were treated with methyl iodide to distinguish C-C linkages from mono-sulfidic linkages (refs. 15-17).

Results and discussion

Crosslink structural analysis

Experimental data for the gum NR compounds show that cure conditions and the use of Pk900 both affect the type of crosslinks that are formed (table 3).

[TABULAR DATA OMITTED]

Pk900 does not influence crosslink structure when the system is at optimum cure, i.e. [t.sub.90]. Total crosslink density and crosslink types of the control and experimental compounds are similar. This supports the claim that Pk900 remains inactive until reversion begins. This is an important consideration that will be revisited when discussing compound physical property data.

The influence of Pk900 becomes apparent when reversion occurs during overcure. The control compound has a 60% reduction in total crosslink density and the crosslink type is 95% mono-sulfidic after overcure. The Pk900 compound also has a reduction in crosslink density, but the reduction is less than that of the control. The decrease in poly-sulfidic crosslinks is compensated by introduction of Pk900 crosslinks. This is clearly evident from the detection of C-C crosslinks in Pk900 samples upon overcure.

In addition to compensating for the loss of crosslinks, the Pk900 improves the compound's heat stability. The Pk900 crosslinks that are formed are more heat stable than the sulfur crosslinks they replace based on bond energies (table 1). The higher bond strength for the C-C bonds is the basis for improved compound heat resistance.

A schematic representation showing the incorporation of Pk900 crosslinks during the reversion process, is shown in figure 4. It should be noticed that Pk900 crosslinks do not begin to be incorporated until reversion begins. This is the basis for Pk900 causing little or no change in scorch, cure rate or optimum properties.

[TABULAR DATA OMITTED]

The results of this analysis show that when Pk900 is used in a compound that is overcured, a new crosslink network is developed. This new network is based on more heat stable crosslinks. Additionally, this crosslink network gives the overcured Pk900 compound physical properties that are very similar to the control compound cured to [t.sub.90]. An additional benefit from the use of Pk900 is that the new carbon-carbon based crosslinks lead to a vulcanizate that is more heat stable in service. Higher temperature service conditions are possible.

Compound physical property data

Cure characteristics

The rheometer cure curves for a conventionally cured natural rubber compound are shown in figures 5 and 6. It is clear from the curves that the control compound exhibits reversion during extended curing. The incorporation of Pk900 into the compound is also demonstrated in figures 5 and 6. As Pk900 crosslinks are incorporated, the compound's maximum torque is maintained. Quantitatively, the 150[degrees]C data shows that the control's maximum torque value drops 18%, while the compound containing Pk900 has no drop in maximum torque.

[ILLUSTRATION OMITTED]

At 170[degrees]C (figure 6), the control compound shows even more reversion. The Pk900 compound shows slight reversion, but the percentage of reversion is still substantially less than the control compound.

An additional feature of the Pk900 is that it has no influence on cure parameters. Since Pk900 does not become active until reversion occurs, scorch and cure time are not affected by its addition. No special compound modifications are needed when Pk900 is added to a compound.

Technological properties

Stress-strain properties - unaged

The effect of Pk900 on stress-strain properties can be seen from table 4. The compound containing PK900 better maintains its modulus on overcure. This is evident at both 150[degrees] and 170[degrees]C. This is a clear sign that Pk900 is compensating for the loss of crosslink density that occurs during thermal decomposition of the sulfur crosslinks.

[TABULAR DATA OMITTED]

It needs to be mentioned here that the Pk900 compensation effect may result in a small increase in modulus on overcure at 170[degrees]C. This increase can be addressed by varying the Pk900 level in the compound.

Use of Pk900 improves the tensile property of the vulcanizate. This is particularly noticeable on overcure, where the loss in crosslink density contributes to the drop in tensile of the control compound.

When evaluating new crosslinks, it is important to address the heat resistance vs. fatigue property compromise. As previously mentioned, reversion can be addressed by modifying the cure system to produce more heat stable mono-sulfidic crosslinks. These crosslinks lack flexibility however, and therefore tend to suffer from poor fatigue properties.

Fatigue testing shows that the compound containing Pk900 has equal or better properties than the control. It is important to note that this is even true on overcure, when the Pk900 based compound has maintained the designed modulus and the control has lower modulus. Generally, lower modulus compounds perform better on fatigue tests. This demonstrates that the Pk900 crosslinks are more flexible than the mono-sulfidic crosslinks that remain in the control after overcure.

Stress-strain properties - aged

The effect of oxidative aging is reflected in the modulus and tensile strength of the vulcanizates. Retention of physical properties, namely modulus and tensile strength, thus provides a measure of age-resistance.

Aged tensile data obtained after aging at 100[degrees]C for 72 hours is shown in figure 7. It can be seen that vulcanizates containing Pk900 have a small, but positive effect on retention of strength properties during aging.

[ILLUSTRATION OMITTED]

During aging, the Pk900 crosslinks are more stable than the conventional sulfidic crosslinks. Unlike sulfidic crosslinks (ref. 18), the crosslinks introduced by Pk900 are thermally stable and highly susceptible to nucleophilic, electrophilic and free radical attack. The Pk900 crosslinks are therefore resistant to further reactions while the rubber compound is in service.

Heat build up characteristics

The greatest enemy of a rubber compound is heat developed during dynamic strain. If allowed to build up unchecked, this heat will eventually cause chemical and physical degradation of the vulcanized rubber and lead to a disastrous loss of strength (ref. 10). The heat development is a result of internal friction resulting from the inevitable flexure of the compound in service. The generation of heat is further related to load under which flexure takes place. Large truck tires, off-the-road tires and some engineered parts are required to operate under severe conditions. Repeated deformation, under high loads, for extended periods of time are not unusual service conditions. This type of service can cause excessive heat generation within a tire or part resulting in a premature failure. Controlling this heat is therefore a real issue that compounders have faced for years.

One of the best test methods to measure the irreversible change to the crosslink system that occurs during the conditions described above is the Goodrich Flexometer (ref. 3). Figures 8-11 show the results of testing the control and experimental compounds at different cure conditions. Although there are small differences in heat build up for optimum cured samples, significant improvement is seen with the overcured samples. The overcure conditions are more realistic cure conditions for thick section rubber articles.

[ILLUSTRATION OMITTED]

The reduction, and more so, stabilization of heat generation is based on the experimental compound maintaining modulus under conditions that cause reversion.

An extended flexometer heat build-up test gives additional information concerning the total hysteresis losses during long term flexing of rubber goods in dynamic application. Figure 12 shows the results of this test. The Pk900 samples attain an equilibrium temperature in less than an hour and maintain that temperature for six hours without failure, while the control samples do not survive one hour. Testing on a semi-efficiently cured, NR/BR compound gave similar results. The control ran for 90 minutes before failure, while the Pk900 sample was run for 18 hours without failure. Figure 13 shows the flexometer pellets after this testing was concluded.

[ILLUSTRATION OMITTED]

Increasing the severity of the test conditions leads to a failure of both the control and the experimental compound. The reduction of heat generation, however, leads to an improvement in the blow-out times of the Perkalink 900 compound sample.

Dynamic properties

The effect of Pk900 on the dynamic properties is illustrated in figures 14-16. The control compound shows a drop in complex modulus arising from reversion and significantly larger changes in tan [delta] and loss compliance than shown by the compound containing Pk900. The reduction of loss compliance suggests that the use of Pk900 will favorably influence the rolling resistance of pneumatic tires. These findings confirm the reversion protection by Pk900 via compensation of loss of crosslink density.

[ILLUSTRATION OMITTED]

Conclusions

Pk900 is a chemical that protects rubber compounds from the process of reversion. Whether the reversion is caused by overcure, high temperature curing or high temperature application, Pk900 compensates for the loss of compound properties leading to improved compound performance.

Pk900 is unique in that reversion is required for it to become active. Pk900, therefore, addresses the problem of reversion without affecting scorch, cure rate or optimum properties. In practice, this means that it can be added without making any other compound or process modifications.

The crosslinks formed by Pk900 are flexible. They do not negatively affect fatigue properties like some other crosslink systems designed to address reversion. Additionally these new crosslinks are more heat stable, which will allow natural rubber compounds to operate in more severe service conditions.

As a consequence of crosslink compensation, compounds using Pk900 will benefit in a variety of ways:

* Maintenance of design properties throughout rubber article.

* Reduced heat generation in service.

* Property maintenance without a loss of fatigue or mechanical properties.

References

(1.)J.I. Cunneen and R.M. Russel, J. Rubber Res. Inst. Malaysia, 22 (3), 300, 1969.

(2.)A.B. Sullivan and R.W. Wise, "Rubber Technology," Third Edition, M. Morton, Ed., NY 1987, Ch. 4.

(3.)A.I. Medalia, Rubber Chem. Technol., 64, 481, 1990.

(4.)S. Wolff, Kautschuk Gummi Kunststoffe, 32, 760, 1979.

(5.)D. Lloyd, Eur. Rubber Journal, 27, 1988.

(6.)Th. Kempermann and V. Eholzer, Kautschuk Gummi Kunststoffe, 34, 722, 1981.

(7.)G.M. Bristow, NR Technology, 17, 7, 1981.

(8.)D. Eemans, Kautschuk Gummi Kunststoffe, 46, 371, 1993.

(9.)G. Horper, K.H. Nordsiek and Z. Zerpner, EP-A-0432417, 1989.

(10.)G.M. Bristow and B.T. Wistall, J. Appl. Polym. Sci., 9, 495, 1965.

(11.)J. Mullins, J. Appl. Polym. Sci., 2, 1, 1959.

(12.)J. Mooney et al., J. Appl. Physics, 11, 100, 1940.

(13.)R.S. Rivlin, Philus. Trans. R. Soc., London, Ser A 243, 251, 1951.

(14.)B. Saville and A.A. Watson et al., Rubber Chemistry Technology, 40, 100, 1967.

(15.)D.S. Campbell et al., J. Appl. Polym. Sci., 13, 1201, 1969.

(16.)G.C. Moore et al., J. Polym. Sci., 19, 237, 1956 and 32, 503, 1958.

(17.)M.L. Selker et al., Ind. Eng. Chem., 36, 20, 1944.