Activation systems for improved aging in epichlorohydrin terpolymer compounds.

One of the unique properties of epichlorohydrin terpolymers is that they contain dual cure sites, including: chlorine contributed by the epichlorohydrin (ECH) monomer and a double bond contributed by the allyl glycidal ether (AGE) monomer, as shown in scheme 1. The chlorine allows for curing with the Zeon Chemicals non lead curative system known as Zisnet (TCY) or an ethylene thiurea (ETU) based system. The unsaturation present due to the AGE monomer enables conventional (sulfur or peroxide) curing. This, of course, necessitates the inclusion of a metal oxide activator in the cure system, which also acts as an acid scavenger to protect against acid catalyzed cleavage of the ether linkage in the polymer backbone, the principal degradation mechanism in polyether elastomers (ref. 1). One of the most efficient and common metal oxides used in curative activation is, of course, zinc oxide.

[ILLUSTRATION OMITTED]

Unfortunately in the case of epichlorohydrin elastomers, zinc-containing materials can readily be converted to zinc chloride during extended aging at or near 125[degrees]C. Zinc chloride is a strong Lewis acid and catalyzes backbone chain scission of the polyether linkages (ref. 2). In their excellent work investigating catalysis of coal liquefaction, Frederick and Bell studied the effects of zinc halides, and specifically zinc chloride on dibenzyl and cycloaliphatic ethers (ref. 3). While the rubber compound is certainly much more complex than their proposed mechanism, shown in scheme 2, the attack on the ether linkage of the dibenzyl ether by zinc chloride, as shown, is a viable option for understanding the degradation mechanism in polyether elastomers.

[ILLUSTRATION OMITTED]

There are other viable metal oxides for activating cure systems, for example, calcium oxide. While calcium oxide is a good choice (it does not form a prodegradant complex), it is not as active as zinc oxide. It, therefore, gives a "lazier" cure curve than zinc oxide, as shown in figure 1. The challenge, then is to determine if a proportion of zinc oxide could somehow be used in combination with calcium oxide to hasten the activation of compounds, with no detrimental effects on aging. In addition to the calcium/zinc system, other options for activation were also evaluated, such as zinc stearate and zinc dimethacrylate (ZDMA-SR708).

[FIGURE 1 OMITTED]

Experimental

A series of carbon black filled, sulfur and peroxide cured, compounds was prepared using a GECO T3100 (terpolymer) as base elastomer. Four metal oxide type compounds were evaluated at various levels and ratios. These are detailed in table 1.

The stocks were one-pass mixed in a BR1600 laboratory internal mixer and sheeted off on a 20.3 cm x 20.3 cm x 40.7 cm laboratory mill with a roll speed ratio of 1.28:1. The mill was prepared by cold cleaning with Glob abrasive stock and then preheating to 60[degrees]C-80[degrees]C for sheeting. After rheology testing on an ODR for 30 minutes at 180[degrees]C, the stocks were molded into standard ASTM 15.2 cm x 15.2 cm tensile sheets in a 60.9 cm x 60.9 cm, 100 ton electric press. Curing was completed for Tc90 at 180[degrees]C.

Stress-strain properties were measured using ASTM D412 procedures before and after aging for 70, 168, 336, 504 and 1,008 hours at 125[degrees]C in a hot air convection oven. Samples were retained from these interval tests and characterized analytically using high-resolution thermogravimetry (hi-res TGA). This technique involves the rapid heating of a sample until a certain rate of weight loss is detected. The instrument's software then slows the heating rate to maintain the rate of heat loss, often to nearly isothermal conditions. After the weight loss falls below a certain level, the rapid heating is resumed until the next weight loss is detected, or at the end of the run. This technique is quite useful, as it tends to sharpen all weight losses except those controlled by diffusion.

Standard calcium and zinc oxide activation

In the first experiment, the levels of zinc and calcium oxides were held constant at 5 phr, a typical loading in sulfur and peroxide cured compounds. Table 2 illustrates the ODR rheology and original physical properties of the compounds. In the sulfur cured stocks, dramatic differences in activation between zinc and calcium oxides can be seen (figure 1). The activation sluggishness is not seen with the peroxide system, which activates quite nicely with the calcium oxide (figure 2). Even though good activation is obtained with calcium oxide in the peroxide system, and zinc oxide is not needed, this type cure system is still included in the study, both for continuity and academic interest. Note also in figure 2 the degradation observed as reversion in the latter portion of the ODR cure curve for the zinc-oxide-containing compound, which is reason enough not to use this activator with a peroxide system.

[FIGURE 2 OMITTED]

Since it is known that the principle mode of ECO degradation is backbone chain scission of the ether linkages, observing changes in tensile strength after aging is most telling. Observation of elongation changes is also important, since these equate with loss of service as well.

After aging the stocks in hot air (125[degrees]C) for 1,008 hours, the results were dramatic. In figures 3 and 4, the zinc oxide activated compounds were destroyed, as shown by severe losses in tensile and elongation. (The -100% value shown is for graphical continuity only. The samples were actually destroyed.) Conversely, the calcium oxide activated compounds retained most of their tensile strength, and, in the case of the peroxide cured stock, 148% actual elongation. That the sulfur stock lost significant elongation is not surprising, given the aging conditions. It is notable, however, that no catastrophic loss of tensile occurred, indicating polymer backbone integrity.

[FIGURES 3-4 OMITTED]

TGA analysis of the unaged vs. 1,008-hour aged compounds confirms the results found by physical testing of the stress-strain properties. In figures 5 and 6, the onset of degradation temperature is substantially lower for the zinc-oxide-containing compounds relative to the original, unaged samples. This indicates polymer loss due to degradation. In figures 7 and 8 ,the calcium-oxide-containing compounds are shown to be unchanged in onset of polymer degradation temperature after 1,008 hours, indicating there is no loss of polymer (degradation).

[FIGURES 5-8 OMITTED]

Varying ratios of calcium and zinc oxides

In a second experiment, the levels (phr) of zinc and calcium oxides were ratioed 4:1 phr (CaO:ZnO) and 1:4 phr (CaO: ZnO), and the ODR rheology as well as original physical properties checked, shown in table 3. The rheological and original stress-strain properties are similar within the sulfur cured and peroxide cured compounds. As noted in figure 9, the higher ratio of zinc oxide activated the stocks to a faster extent than did the calcium.

[FIGURE 9 OMITTED]

Figures 10 and 11 show that after 1,008 hours at 125[degrees]C, the sulfur and peroxide cured compounds containing the 4:1 phr ratio (CaO:ZnO) had very minimal tensile losses (21%). As before, the elongation losses of the sulfur cured compound were large, as would be expected. The aged 1:4 phr ratio (CaO:ZnO) compounds (figures 10 and 11) had tensile responses that revealed significant polymer degradation in the sulfur cured compound. Interestingly enough, this was not the case in the peroxide cured compound, where tensile and elongation losses were moderate. This is especially intriguing in light of the destruction found with the 5 phr zinc oxide, peroxide cured stock shown earlier (figure 6).

[FIGURES 10-11 OMITTED]

The TGA results for the 1:4 phr ratio (CaO:ZnO) compounds are shown in figures 12 and 13. As expected, the sulfur cured compound had a lower polymer degradation temperature after aging. The peroxide cured version showed an anomaly not readily explainable; the polymer degradation temperature was actually higher after aging than before. This was reproduced several times. This confirms the physical test results of minimal tensile and elongation changes. Both the 4:1 phr ratio (CaO:ZnO) compounds, which showed very little loss in tensile, also had no change in polymer degradation temperature, as noted in figures 14 and 15.

[FIGURES 12-15 OMITTED]

Considering the previous results, additional compounds were prepared as before to examine a middle ground of sorts, using a ratio (phr) of 3:2 (CaO:ZnO). The rheological and original physical properties for these compounds are shown in table 4. The activation as noted by the ODR cure curve is quite good (figure 16). After aging for 1,008 hours at 125[degrees]C, both compounds showed excellent responses in terms of both tensile and elongation (figures 17 and 18). The TGA plots confirmed this observation with both the unaged and aged compounds showing essentially unchanged polymer degradation temperatures, as shown in figures 19 and 20.

[FIGURES 16-20 OMITTED]

Activation with zinc complexes

In a third experiment, a series of stocks was evaluated where the activation was accomplished using zinc stearate at 5 phr or zinc dimethacrylate (ZDMA) at 15 phr. The ODR theology as well as original physical properties are shown in table 5. The best activation occurred using zinc stearate in the sulfur cured compound. The peroxide cured compound did not activate as sharply. The ZDMA was only evaluated in a sulfur cure system and did appear to show some reversion (degradation) in the latter portion of the cure curve. Figure 21 is illustrative of these points.

[FIGURE 21 OMITTED]

After aging the zinc stearate-containing compounds for 1,008 hours at 125[degrees]C, a steady decline in tensile and elongation was observed (figures 22 and 23). The changes were especially severe in the peroxide-containing compound. These results seem to indicate that zinc stearate is not viable in epichlorohydrin terpolymer compounds in higher temperature service.

[FIGURES 22-23 OMITTED]

Interestingly enough though, the TGA testing of these compounds revealed essentially no change in polymer degradation temperatures of the 1,008-hour agings relative to the original, unaged specimens (figures 24 and 25). This could indicate that the change in physical properties is related strictly to thermal-oxidative aging phenomena and not polymer chain scission, which of course in the end could lead to the same condition.

[FIGURES 24-25 OMITTED]

The ZDMA sample showed declines in tensile and elongation, but not to the same extent as the zinc-stearate-containing compounds (figures 22 and 23). The actual elongation remaining was 171%. TGA results comparing the unaged and aged specimens indicated that essentially no polymer degradation is occurring. However, as with zinc stearate, the most probable degradation mechanism is thermal oxidative.

Conclusions

While zinc oxide is a widely used activator in conventional (sulfur and peroxide) cure systems, and provides fast activation, it is not recommended for use with polyepichlorohydrin elastomers where higher temperature (at or near 125[degrees]C) services are anticipated. This is because it can complex with chlorine molecules abstracted from the polymer chain and form the prodegradant, zinc chloride. Calcium oxide can also complex with chlorine molecules, but the complex is not a prodegradant and no detrimental effects are found after aging. However, it does not activate as quickly as zinc oxide.

Blends as high as 3 phr calcium oxide and 2 phr zinc oxide are found to activate terpolymer compounds effectively and possess excellent aging stability. The use of zinc containing materials such as zinc stearate in a sulfur cure system is found to be adequate in cure activation. Zinc stearate in a peroxide cure system and ZDMA with sulfur show lazier cure curves. All are questionable in aging performance.

While they do not appear to undergo chain scission, they appear to be degraded by thermal oxidative processes.

This article is based on a paper presented at the technical meeting of the Rubber Division, ACS, October 10-12, 2006.

References

(1.) Kathy Owens and Vern Kyllingstad, "Synthetic polymer elastomers", Kirk-Othmer Encyclopedia of Chemical Technology--Fourth Edition, Vol. No. 8, p.1,087.

(2.) ibid, p. 1,086.

(3.) Tim J. Frederick and Alexis T. Bell, "Cleavage of dibenzyl ether in the presence of zinc halides," Journal of Catalysis--87, pp. 226-237 (1984).

by Samuel C. Harber, Zeon Chemicals ([e-mail])

Table 1 - evaluation recipes

                                             Sulfur    Peroxide

GECO T3100                                    100        100
N550 carbon black                              50         50
Alkyl alkylether diester adipate               10         10
4,4'-bis(a,a-dimethylbenzyl)diphenylamine       1          1
2-mercaptotoluimidazole                         0.5        0.5
Stearic acid                                    1          1
Sulfur                                          0.8
TMTD                                            1
N-(cyclohexylthio)phthalimide                  0.25
a,a-bis(t-butylperoxy)diisopropyl benzene                  2
                                             Varied    Varied
Zinc oxide                                     1-5        1-5
Calcium oxide                                  1-5        1-5
Zinc stearate                                  3          3
ZDMA (Sartomer SR-708)                        15          0

Table 2 - calcium oxide versus zinc oxide; rheology and original
physicals

                            Sulfur                Peroxide
                       CaO (5)   ZnO (5)    CaO (5)   ZnO (5)
ODR 180[degrees]C
ML (dN x m)              11.4       13.7      14.0       13.3
MH (dN x m)              88.0      101.7      84.0       61.0
Ts2 (min.)                2.2        1.2       1.1        1.0
Tc90 (min.)              24.1       14.7       8.4        4.1
Original physicals
Hardness A, points       69         71        61         55
M100, MPa                5.4         5.4       2.9        2.2
Tensile, MPa            17.0        16.2      12.3       10.9
Elongation, %          323        320        431        562

Table 3 - rheology and original physicals
for low and high levels of calcium and
zinc oxides

                        Sulfur (CaO/Zn0)      Peroxide(Ca0/2n0)
                         4/1        1/4        4/1         1/4
ODR 180[degrees]C
ML (dN x m)             11.6       13.5       14.1         13.8
MH (dN x m)            117.1      109.2       90.0         83.3
Ts2 (min.)               1.6        1.2        1.0          1.0
Tc90 (min.)             21.1       19.0        7.5          8.0
Original physicals
Hardness A, points      75         72         62           57
M100, MPa                7.0        6.1        3.5          2.4
Tensile, MPa            16.3       17.3       12.4         11.9
Elongation, %            251        274        396          480

Table 4 - rheology and original physicals for calcium and zinc oxides
at 3/2 phr

                          Sulfur (CaO/ZnO)    Peroxide(Ca0/Zn0)
                                   3/2                 312
ODR 180[degrees]C
ML (dN x m)                       13.7                14.5
MH (dN x m)                      108.7                85.1
Ts2 (min)                          1.4                 1.1
Tc90 (min)                        19.5                 8.3
Original physicals
Hardness A, points                71                  59
M100, MPa                          5.9                 2.5
Tensile, MPa                      17.1                11.9
Elongation, %                    295                 448

Table 5 - rheology and original physicals
of zinc complex materials

                               Sulfur             Peroxide
ODR 180[degrees]C           Zinc     ZDMA             Zinc
                      stearate (5)   (15)         stearate (5)

ML (dN x m)                  11.6    11.0         12.0
MH (dN x m)                  81.5    43.2         40.8
Ts2 (min.)                    1.4     1.0          1.2
Tc90 (min.)                  15.7    22.4          3.5
Original physicals
Hardness A, points           65      55           48
M100, MPa                     4.0     2.1          1.4
Tensile, MPa                 16.4    13.9         10.3
Elongation, %               393     668          691