The effect of HTS on cure kinetics of accelerated sulfur vulcanization.

Understanding anaerobic thermal aging (more commonly known as reversion) began with the elucidation of vulcanizate crosslink structures (ref. (1)). Later, studies of model systems gave more detailed insight into the chemistry of vulcanization and thus reversion (ref. (2)). The main changes occurring during reversion are summarized as follows:

* The total number of crosslinks decreases, resulting in a decrease in modulus;

* The average length (number of sulfur atoms) in a crosslink decreases;

* The amount of monosulfidic bonds increases;

* The rubber chain becomes chemically modified, increasing amounts of cyclic sulfide, conjugated trienes and cistrans isomerization of double bonds; and

* The concentration of extra-network material, such as zinc sulfide increases.

  Table 1 - NR masterbatch

SMR CV-60            100.00
N-330 carbon black    50.00
Flexon 580             5.00
Zinc oxide             5.00
Stearic acid           2.00

Total                162.00

During reversion several reactions occur. These reactions include desulfurization. The chemistry, which ultimately forms polysulfide crosslinks, continues until the capacity to donate sulfur is depleted. In a conventional formulation, at low temperatures and optimal cure times, the initially formed network is rich in polysulfidic bonds (ref. (3)). Continued heating, or cure at higher temperatures, precipitates a reduction of the number of polysulfidic crosslinks, disulfidic crosslinks, and the total number of crosslinks. In addition, the number of monosulfidic crosslinks and, most importantly, chain modifications and zinc sulfide concentrations increase.

Efficient or semi-efficient vulcanization systems are low in sulfur content and high in accelerator. These systems produce networks rich in monosulfidic and disulfidic crosslinks. Consequently, the degree of reversion observed is lower than that observed in conventional formulations. However, vulcanizates produced using these systems often lack the high tear strength and excellent fatigue resistance characteristic of vulcanizates cured using conventional formulations (ref. (4)).

Studies (ref. (5)) using model crosslinks have shown the effects of various proposed vulcanization intermediates on desulfurization chemistry. Additives were heated to 140[degrees]C for two hours with methylpentenyl trisulfides. Desulfurization was assessed afterwards by the determination of the mole percentage of products. Adding high levels of zinc oxide had only a small effect on desulfurization. Also, adding zinc benzothiazole-2-thiolate had only a small effect on desulfurization. This was attributed to the low solubility of zinc benzothiazole-2-thiolate. Adding zinc accelerator complexes, ligated with amines, was particularly effective at desulfurating the model crosslinks. Also, the addition of extra ligands or zinc complexes, such as cylcohexylamine and zinc propionate, further enhanced desulfurization. Thus, the reagents responsible for crosslinking are likely to be the culprits responsible for reversion.

[ILLUSTRATION OMITTED]

[ILLUSTRATION OMITTED]

The use of organic thiosulfates as stabilizing agents for rubber vulcanizates was introduced in the early 1980s (ref. (6)). Early results showed that using alkylene bis(thiosulfates) dramatically improved physical properties, both in optimally cured and over-cured compounds. In the original European patent application, disodium decamethylenebis(thiosulfate) was used as an example for a postvulcanization stabilizer. Upon curing at 140[degrees]C for 200 minutes, the stabilized compound had over 98% retained modulus with only a 20% reduction in fatigue-to-failure resistance. By comparison, an unstabilized control compound retained only 78% modulus and lost 35% of its fatigue-to-failure resistance.

[ILLUSTRATION OMITTED]

[ILLUSTRATION OMITTED]

[ILLUSTRATION OMITTED]

Bis(thiosulfates) have also been shown to be beneficial for other properties as well. In rubber-to-metal composites (ref. (7)), bis(thiosulfates) often improve the initial adhesion strength. In general, adhesion strengths of rubber-to-metal composites aged under hot air, steam or salt water or overcured are much improved compared to a control compound without the bis(thiosulfates). Dynamic and physical properties are also improved in rubber blends after thermal again, even in the absence of transition metal ions such as cobalt (ref. (8)). Bis(thiosulfates) provide improved overall adhesion characteristics with excellent retention of bond strength in saline environment. These properties are observed in steel skim compounds without the use of silica and at stearic acid levels of 2 phr (ref. (9)).

The beneficial effects have been attributed to the formation of hybrid crosslinks in the ensuing network (ref. (10)). This conclusion is based on model studies using cylcohexene as the model for rubber and sodium butyl thiosulfate as the model compound for the post vulcanization stabilizer. These two reagents were mixed with a sulfur curing system, sealed in a vacuum tube and heated to 140[degrees]C for one hour. The product mixture was analyzed by gas chromatography. In the absence of the butyl thiosulfate, di(3-cyclohexenyl) disulfide and di(3-cyclohexenyl) trisulfide were formed as major products. With butyl thiosulfate present, 3-cyclohexenyl butyl polysulfides were present in large amounts together with butyl sulfides. The asymmetric disulfide compounds were identified by mass spectroscopy and by comparison of GC retention times with independently synthesized compounds. Other experiments, performed with 2-methyl-2-pentene as a model compound for natural rubber, provided similar results.

In conventional formulations, polysulfidic crosslinks constitute the majority of crosslinks formed in vulcanizates prepared with an without bis(thiosulfates) salts (ref. (11)). Conventional formulations with post vulcanization stabilizers exhibit low levels of reversion. Upon over-cure, vulcanizates prepared in the presence of the bis(thiosulfate) salts, retained about 60% of the polysulfidic crosslinks. In contrast, the control compound retained only 35% polysulfidic crosslinks. Various measures of cure efficiency and average number of sulfur atoms per crosslink upon over-cure in the bis(thiosulfate) containing compounds. Thus, the active complex formed during vulcanization in the presence of a bis(thiosulfate) is less efficient at desulfurization than that formed in the conventional system.

Recent results, using a new halobutyl elastomer, have demonstrated that disodium bis(hexamethylene thiosulfate) is competitive with a conventional sulfenamide accelerator as a ligand for zinc ion (ref. (12)). Accumulated results of all of these experiments will be used to infer information regarding the active complex of vulcanization modified with bis(thiosulfate) salts.

Experimental work

Masterbatch

Rubber was mixed with carbon black and oils in a "00" internal mixer. The formulation for the masterbatch is given in table 1. The sulfur, zinc oxide and accelerators were added on a two roll mill.

Rubber laboratory evaluation

Vulcanization characteristics were determined using either the oscillating disc rheometer, Monsanto ODR2000, or the moving die rheometer, Monsanto MDR2000. Viscosity, scorch, rheometry and tensile testing were performed in accordance with the appropriate ASTM procedures.

Data treatment

Viscosity and scorch data

The standard test method for viscosity and vulcanization characteristics by use of the Mooney viscometer used was ASTM D1646-93. This method measures the viscosity of a compound which has been heated to and maintained at a specific temperature. The times observed for the viscosity to change by specified amounts are reported.

Initial viscosity, viscosity of the compound measured at the start of the test.

Minimum viscosity, lowest recorded viscosity observed during the test.

t2, time required for the viscosity to increase two Mooney units above the minimum observed viscosity.

t5, time required for the viscosity to increase five Mooney units above the minimum observed viscosity.

t15, time required for the viscosity to increase 15 Mooney units above the minimum observed viscosity.

t35, time required for the viscosity to increase 35 Mooney units above the minimum observed viscosity.

Rheometer data

The definition of vulcanization characteristic parameters, rheometer designs and testing methods can be found in the 1993 Annual Book of ASTM Standards Volume 09.01, Section 9 "Rubber, natural and synthetic - general test methods; carbon black." All of the tests and instruments were calibrated according to ASTM D2084-93. Most of the value are generally reported according to ASTM D2084-93. A summary of the definitions used are included here for the readers' convenience. Deviations from the standard definitions or analogous standard definitions are noted accordingly.

Rmax, dNm, the maximum in rheometer torque, (same as [M.sub.H] in D2084-93);

Rmin, dNm, the minimum in the rheometer torque, (same as [M.sub.L] in D2084-93);

t2, min., the time required for the torque to rise two dNm above the minimum torque, (designated as [t.sub.S]2 in D2084-93);

t25, min., the time required to reach 25% of the torque increase or minutes to reach [M.sub.L] + [0.25(M.sub.H] - [M.sub.L]), (designated as t'25 in D2084-93);

t50, min., the time required to reach 50% of the torque increase or minutes to reach [M.sub.L] + [0.50(M.sub.H] - [M.sub.L]), (designated as t'50 in D2084-93);

t90, min., the time required to reach 90% of the torque increase or minutes to reach [M.sub.L] + 0.90 [(M.sub.H] - [M.sub.L]), (designated as t'90 in D2084-93).

[ILLUSTRATION OMITTED]

Tensile properties

The standard test method for vulcanized rubber and thermoplastic elastomers - tension, designated D-412-92 was followed in measuring the tensile properties. Die-C dumbbell specimens (method A) were used in testing. Reported values include the following:

TS, MPa, tensile strength, the stress at rupture;

E, %, elongation (at rupture) in percent of the original sample length;

[T.sub.1.sub.0.sub.0], MPa, tensile stress at 100% elongation;

[T.sub.2.sub.0.sub.0], MPa, tensile stress at 200% elongation; and

[T.sub.3.sub.0.sub.0], MPa, tensile stress at 300% elongation.

Tear strength

The standard test method for trouser tear strength of conventional vulcanized rubber and thermoplastic elastomer, designated ASTM method D624-91, was followed in measuring the tear strength of the samples. The mean tear strength, [T.sub.s], is reported in Newtons.

Results and discussion

Hexamethylene bis(thiosulfate) disodium salt, (HTS), was used as the post vulcanization stabilizer. It was compounded into a natural rubber masterbatch (table 1) with a conventional formulation having 2.5 phr sulfur, 0.6 phr 2-mercaptobenzothiazole t-butyl sulfenamide (TBBS) and 5.0 phr zinc oxide. Mole ratios of HTS to TBBS range from zero to about three and a half. Table 2 contains formulations, viscometer, rheometer, tensile and tear data for the range of compounds tested.

Figure 1 shows the Mooney scorch delay at 121[degrees]C. Scorch delay is a slightly decreasing function of HTS loading. Less than 20% of the scorch delay is lost at loading up to 3.5 moles HTS per mole TBBS.

Figure 2 shows the moving die rheometer ts2 data for cure data collected at 150[degrees]C and 160[degrees]C. The 150[degrees]C data show a decrease in the value of ts2 at with increasing loading of HTS. The decrease is most notable between the loadings of zero and two moles HTS per mole TBBS. About a third of the scorch delay is lost at the 3.5 loading of HTS. The 160[degrees]C data for ts2 also decreases, but the magnitude of the decrease is less than that observed at 150[degrees]C.

Time to reach optimal cure, t90, increases with increasing levels of HTS. At either 150[degrees]C or 160[degrees]C, little change is observed in the range of 0 to 1 moles HTS per mole TBBS. Above the one mole HTS loading level the t90 times increase linearly with HTS loading at both cure temperatures.

[ILLUSTRATION OMITTED]

The extent of cure can be estimated by rheometer Mh. Figure 3 shows a small reduction in Mh, about 10%, over the range of HTS tested. The maximum rate of vulcanization decreases in a continuous fashion with increasing HTS levels. Both of these properties appear to change monotonically with HTS levels.

The reversion resistance was estimated by a loss in rheometer torque. The difference between maximum torque and the torque observed at a time of five times t90 was calculated. The calculations were done for both rheometer cure curves measured at 150[degrees]C and 160[degrees]C. Figure 4 shows that this difference is monotonically reduced as the mole ratios of HTS to TBBS increases. The maximum difference of greater than 3.5 dNm, for the data point at zero loading HTS and 160[degrees]C, represents about a 25% loss in torque from the maximum torque.

While most of the data noted above show monotonic decreases or increases in properties with respect to HTS loading, some curves show a distinct change in slope at the 1.0 mole ratio of HTS to TBBS. This is noticeable in the Mooney scorch data and the rheometer t90. The activation energy of cure was determined using the moving die rheometer for a range of 0.0-2.5 moles HTS to TBBS. Figure 5 shows that the activation energy is a nonlinear function of HTS loading. Almost no change is observed for HTS loading below one mole HTS per mole TBBS. At greater than one mole ratios, the activation energy increases with increasing loading of HTS.

The tensile moduli, given here as a measure of extent of cure or effective crosslink density, are compared in figure 6. When cured to t90 or short over-cure, 3 x t90, there is functionally no dependence in the tensile properties on HTS loading. Trouser tear test data are summarized in figure 7. At room temperature or at 100[degrees]C, no change in tear is measured for low level of HTS. However, when higher loadings of HTS are used (greater than one mole HTS per mole of TBBS) significant increases in tear strength are realized. In the hot tear test, nearly 50% improvement in tear strength is measured at the 2.5 mole ratio HTS loading.

Conclusions

HTS has been shown to slightly reduce scorch time. Cure times, t90, increase with increased loading of HTS. Little change is observed in the extent of cure as estimated either by the rheometer maximum torque or through tensile modulus values. Reversion resistance increases with increasing HTS loading. The increase in cure activation energy is shown to be related to the stoichiometry of the compound.

The changes in physical properties and cure kinetics, and the improved reversion resistance suggest that either the mechanism of sulfuration has changed, or the nature of the active sulfurating species is unique when HTS is employed. These effects are realized to a greater extent at higher loadings of HTS.

Improved reversion resistance, network sulfur rank analysis on over-cure, and the stoichiometric dependence of activation energy on HTS level support the conclusion that the active sulfurating species is altered. The stoichiometric dependence suggests that the active sulfurating species is likely to contain one or more thiosulfate moieties derived from the HTS. This active species produces crosslinks at a lower rate than conventional sulfenamide derived complexes. The decrease in rate is reflected in the increased activation energy of cure. The increased activation energy of cure is apparently extended to the desulfurization reactions of the crosslinks, thereby resulting in lower rates of reversion.

Other research (ref. (13)) indirectly supports these conclusions; HTS promotes the availability of zinc ion for cure in a novel halobutyl elastomer. In the same work, it was shown that HTS can compete effectively with TBBS for available zinc. The Lewis acidity of HTS coupled zinc ion is judged to be greater than that of sulfenamide ligated zinc ion (ref. (14)). These data support the presence of HTS in the activated complex.

Improvements in tear strength at equal modulus are consistent with formation of a bimodal network (ref. (15)). Sharaf and Mark propose the excellent tear resistant properties observed in sulfur cure elastomers is a result of a population of long chain polysulfidic crosslinks which behave as a bimodal network (ref. (16)). The chemical evidence of the hybrid crosslink (ref. (10)) and the improved tear strength at equal modulus are strong evidence supporting the formation of a bimodal network.

[Part 1 of 2]

                              Table 2

                                     1        2        3        4

A-1 masterbatch                    162.00   162.00   162.00   162.00
Sulfur                               2.50     2.50     2.50     2.50
Santocure-NS                         0.60     0.60     0.60     0.60
ZnO                                  5.00     5.00     5.00     5.00
Duralink HTS                         3.50     2.50     1.50     1.00

Mooney Scorch @ 121[degrees]C
Minimum viscosity                   24.09    24.35     25.2    24.22
t2                                  23.74    26.82    29.41    31.36
t5                                  26.37    29.47    32.39    34.51
t15                                 29.83     32.6     35.5    37.67
t35                                  33.3    34.92    37.46    39.46

MDR2000 @ 150[degrees]C
Rmin                                 1.33     1.31     1.34      1.4
Rmax                                15.74    15.84    16.44    16.82
Rfinal                              15.73    15.81     15.9     15.8
ts1                                  2.11     2.45     3.04     3.37
ts2                                  2.93     3.15     3.63     3.95
t10                                  2.48     2.77     3.37     3.73
t25                                  4.16     4.26     4.51     4.77
t50                                  8.12     7.58     6.92     6.87
t90                                 24.27    20.24    15.33    13.73
Vmax                                  1.5      1.6      2.2      2.5
Tangent delta                       0.067    0.065    0.063    0.064
Rmax-Rfinal                          0.01     0.03     0.54     1.02

MDR2000 @ 160[degrees]C
Rmin                                 1.23     1.27     1.27     1.32
Rmax                                14.11    14.35    14.87    15.34
Rfinal                               13.7    13.58    13.27    13.05
ts1                                   1.4     1.55     1.76     1.85
ts2                                  1.82     1.93     2.13     2.23
t10                                  1.53     1.69     1.92     2.03
t25                                  2.33     2.37     2.54     2.63
t50                                  4.29     3.97      3.8     3.79
t90                                 11.79     9.94     8.09     7.41
Vmax                                  2.5       3.      3.5      3.8
Tangent delta                        0.07    0.067    0.064    0.062
Rmax-Rfinal                          0.41     0.77      1.6     2.29

Tear properties @ 23[degrees]C
Mean stress                         16.14    18.81    14.17    13.62
Std. dev.                            2.87     2.92     1.82     1.17

Trouser tear at 100[degrees]C
Mean stress                            27    27.99    23.11    21.34
Std. Dev.                            1.75     0.15     3.81     0.67

Tensile properties cured to t90
+ 1 @ 160[degrees]C
Cure time                              12       10        8        7
Hardness                               59       56       59       57
T                                   22.05    23.86    25.46     25.9
E                                     507      523      549      543
t100                                 2.34     2.31     2.37     2.26
t200                                 6.01     6.03     6.22     6.09
t300                                10.94    11.16    11.49    11.55
Average t300                        10.84    11.06    11.42    11.58

Tensile properties cured to 3 x
t90 + 1 @ 160[degrees]C
Cure time                              36       30       24       21
Hardness                               62       59       60       61
T                                   21.93    22.62    23.67    23.69
E                                     451      465      481      481
t100                                 2.83     2.74     2.76     2.62
t200                                 7.19     6.93     7.07     6.84
t300                                12.93     12.5    12.73    12.53

[Part 2 of 2]

                              Table 2

                                     5        6        7

A-1 masterbatch                    162.00   162.00   162.00
Sulfur                               2.50     2.50     2.50
Santocure-NS                         0.60     0.60     0.60
ZnO                                  5.00     5.00     5.00
Duralink HTS                         0.50     2.00     0.00

Mooney Scorch @ 121[degrees]C
Minimum viscosity                   24.72    24.16    29.78
t2                                  32.56       28    31.49
t5                                  35.97    30.97    34.96
t15                                 39.11    34.02    38.11
t35                                 40.77    36.15     39.7

MDR2000 @ 150[degrees]C
Rmin                                 1.42     1.34     1.71
Rmax                                17.17    16.08    17.74
Rfinal                              15.48    15.85    15.31
ts1                                  3.75     2.71     4.08
ts2                                  4.39     3.33     4.76
t10                                  4.18     3.03     4.57
t25                                  5.18     4.29     5.49
t50                                  7.02     7.05     7.11
t90                                 12.87    17.27    12.37
Vmax                                  2.6      1.9      2.9
Tangent delta                       0.064    0.066    0.061
Rmax-Rfinal                          1.69     0.23     2.43

MDR2000 @ 160[degrees]C
Rmin                                 1.33     1.27     1.63
Rmax                                15.55     14.7    16.19
Rfinal                              12.49    13.57    12.25
ts1                                  1.93     1.66     1.98
ts2                                  2.36     2.04     2.42
t10                                  2.16     1.81     2.23
t25                                  2.76     2.46     2.81
t50                                  3.77     3.88      3.7
t90                                   6.9     8.98      6.5
Vmax                                  4.1      3.2      4.5
Tangent delta                       0.061    0.065    0.059
Rmax-Rfinal                          3.06     1.13     3.94

Tear properties @ 23[degrees]C
Mean stress                         11.82    17.19    14.61
Std. dev.                            0.56     1.31     0.27

Trouser tear at 100[degrees]C
Mean stress                         20.96    24.58    19.11
Std. Dev.                            5.28     5.84     0.42

Tensile properties cured to t90
+ 1 @ 160[degrees]C
Cure time                               7        9        7
Hardness                               59       56       60
T                                    26.4    24.48    26.92
E                                     553      551      544
t100                                 2.26     2.15      2.3
t200                                 6.09     5.61     6.21
t300                                11.71    10.51     11.8
Average t300                        11.69    10.56     11.9

Tensile properties cured to 3 x
t90 + 1 @ 160[degrees]C
Cure time                              21       27       21
Hardness                               59       59       62
T                                   24.07    23.55    24.78
E                                     477      490      499
t100                                 2.82     2.78     2.62
t200                                 7.43     6.96     6.95
t300                                13.25    12.41    12.72

References

(1) . B. Saville, A.A. Watson, Rubber Chem. and Technol. 40, 100 (1967).

(2) . Bateman, L. Ed., "The chemistry and physics of rubber-like substances," John Wiley and Sons, New York.

(3) . C.T. Loo, Polymer 15, 357 (1974).

(4) . See reference 3.

(5) . Watson, A.A. "The chemistry of vulcanization, of olefinic rubbers by means of tetramethylthiuram disulfide zinc oxide and related systems," Ph.D. Thesis University of London, p. 53 (1965); Porter, M., "Perspectives in the organic chemistry of sulfur," Zwanenburg, B. and Klunder, A.J.H., eds., Elsevier Science, Amsterdam, p. 267 (1987); Roberts, A.D., Ed. "Natural rubber science and technology," Oxford Science Publications, pp. 570-575 (1988), Porter, M. "The chemistry of sulfides," Plenum Press, New York, Chapter 3 (1977).

(6) . Moniotte, Philippe Gerard, Eur. Pat. Appl. EP 70143.

(7) . Delseth, Jean Marc, Mauer, Daniel Emile, Moniotte, Philippe Gerard, Eur. Pat. Appl. 109955.

(8) . Anthione, G., Lloyd, D.G. Ind. Gomma, 29 (8-9), 28-9, 32-3, 36-7, 74, (1985).

(9) . A. Orband, G. Anthione, and H. Roebuck, Kautschuk Und Gummi Kunststoffe 39 (1), 37 (1986).

(10) . G. Anthione, E.R. Lynch, D.E. Mauer and P.G. Moniotte, "A new concept to stabilize cured NR properties, during thermal aging and improve adhesion to brass," paper no. 51, Rubber Division, ACS, (1984).

(11) . See reference 9.

(12) . A.J. Dias and A.A. Galuska, "Curative migration in Exxpro blends," paper no. 96, Rubber Division, ACS, October 1995 and F. Ignatz-Hoover and N. Newman, "Black tire sidewall compounds based on brominated poly(isobutylene-co-p-methylstyrene) elastomer and cured with 1,6-hexamethylene-bis(thiosulfate), paper no. 110, Rubber Division, ACS, October 1995.

(13) . F. Ignatz-Hoover and N. Newman, paper no. 110, Rubber Division, ACS, October 1995 and J.A. Dias and A.A. Galuska, paper no. 96, Rubber Division, ACS, October 1995.

(14) . F. Ignatz-Hoover and N. Newman, paper no. 110, Rubber Division, ACS, October 1995.

(15) . Yanyo, L.C. and Kelly, F.N., Rubber Chem. and Technol., 60, 78 (1987), Smith, T.L., Haidar, B. and Hedrick, J.L., Rubber Chem. and Technol., 63, 256 (1990).

(16) . Sharaf, M.A. and Mark, J.E., J. Macrolmol. Sci., Macromol. Rep., A28, 67 (1991) and Mark, J.E., Accounts in chemical research, 27, 271 (1994).