EPDM extrusions resistant to 175 degrees C.

As a consequence of the constant drive to increase automotive fuel efficiency, underhood service temperatures have risen substantially over the last 20 years. With further environmental pressure to increase CAFE standards, this trend is expected to continue. As a result, the temperature aging requirements of many underhood rubber components have changed substantially.

EPDM polymers, by virtue of their inherent heat resistance and relative low cost, are used in several underhood applications including heater and coolant hoses, air ducting and AC hose covers. Changes in performance requirements, however, have begun to test the upper temperature limits of these EPDM based formulations. In coolant hose, for example, aging resistance requirements have risen to 165[degrees]C. This represents a 40[degrees]C increase over requirements in effect only 15 years ago. General Motors GM6250 radiator hose specification, for example, requires 70% tensile strength retention and 45% elongation retention after one week aging at 165[degrees]C. A maximum 60% compression set after 70 hours at 165[degrees]C is also required. Several original equipment manufacturers have stated a desire to increase requirements to the same levels of performance at 175[degrees]C. Processability problems and cost have, however, prohibited widespread use of EPDM rubber to meet these anticipated requirements.

Formulations meeting the previously mentioned targets at 175[degrees]C were developed by Dunn (ref. 1) in 1987. Dunn found these targets were achievable by careful selection of antioxidants, minimizing plasticizer levels and minimizing base polymer unsaturation. However, these formulations, mineral filled compounds containing 40% to 50% rubber, had impractical processing characteristics for most extrusion applications.

Further work by Dunn and co-workers (ref. 2) was initiated to develop formulations with improved processability. In this work, additives, filler, coagents and peroxide curative levels were varied in statistically designed experiments to improve aging resistance in lower rubber content formulations. As a result, processability was improved. Formulations with between 30 and 35% rubber were developed. Aging targets were met only marginally in their optimized recipe.

In both works, it was concluded that ENB unsaturation levels needed to be below 2.5% to meet aged tensile and elongation retention requirements. Compression set resistance, however, was predictably poorer. Antioxidants, pdicumyl diphenylamine (DCPA) and zinc methyl-mercaptobenzimidazole (ZMMBI), were identified as additives to enhance aging resistance without significant peroxide cure interference.

The use of antioxidants to enhance EPDM aging resistance while minimizing interference with curatives has been studied extensively. Reinvestigations of AO studies may serve to further improve performance beyond this work. In 1980, Ashworth and Hill (ref. 3) reported synergistic effects of 2,2,4-trimethyl-1,2-dihydroquinoline (TMQ) and methyl mercaptobenzimidizole (MMBI). A 1:4 TMQ/MMBI ratio was shown to have better aging performance than either component separately at constant antioxidant concentration. Benefits of polymer bound antioxidant were explored by Scott and Tavokoli (ref. 4). Bound diphenyl amines had better aging resistance than TMQ and other antioxidants, and had the benefit of non-extractability (some peroxide interference problems were noted).

The main purpose of this study was to make improvements to the optimized formulations presented in the previous works to meet the emerging industry need for low cost, 175[degrees]C resistant extrusions with practical processing characteristics. Further examination of base polymer effects was conducted. These included molecular weight and ethylene content variations. In addition, further study of additives was conducted in a designed experiment to improve aging resistance and/or processability.

Experimental

This investigation was conducted in two stages:

* Investigation of base polymer effects in an optimized recipe.

* Designed experiment investigating additive effects.

All formulations were prepared in a 1600 cc internal mixer. A two stage, upside down mixing sequence was used for each formulation with the exception of the metal oxide and antioxidant screening. These compounds were made by mill mixing curatives and additives with aliquots of a blended masterbatch. In the other experiments, curatives were added in a second internal mix sequence.

All testing and aging of samples were completed in accordance with ASTM standards. Molded compression set buttons were used rather than plied tensile sheets.

The effects of base polymer molecular weight and composition were studied in an optimized formulation similar to that generated by Dunn and co-workers (ref. 2). Varying compositions were arrived at by using the five low unsaturation polymers described in table 1. As shown, four of the polymers used are close in design to represent a [2.sup.2] factorial designed experiment with ethylene content and Mooney viscosity as variables.

Table 1 - polymers studied

Polymer ID      Mooney                 %       %
             ML 1+4 (125[degrees]C)   ENB   Ethylene
EPDM A               55               1.0      58
EPDM B               17               3.0      75
EPDM C               50               1.5      66
EPDM D               17               2.0      55
EPDM E               50               1.5      62

The base formulation used for the polymer blend studies is given in table 2.

Table 2 - test formulation for base polymer study

                                      Phr
Polymer                             100.0
N762 black                          100.0
Talc, vinyl silane treated           40.0
DCPA                                  2.0
ZMMBI                                 2.0
Paraffin oil                         30.0
Paraffin wax                          3.0
MgO                                  10.0
ZnO                                   5.0
Stearic acid                          1.0
Dicumylperoxide, 40% active           7.0
Trimethylolpropane trimethacrylate    2.0

Total plasticizer level was lowered further than previous studies. N762 black was used in favor of N650 black to maintain processability. Thirteen separate formulations were derived from base polymers in order to vary base polymer composition. Weighted averages were used to calculate polymer composition. The empirical blending law was used to calculate base polymer viscosities presented along with test data in table 3.

[TABULAR DATA OMITTED]

In figure 1, a comparison of compound viscosity with the calculated base polymer viscosity is shown. The strong correlation supports the use of the empirical blending law to estimate polymer viscosities.

[CHART OMITTED]

Up to this point, several of the formulations presented met the aging targets initially established with good extrudability. Recipe #38, for example, had 45% compression set. Tensile and elongation retention at 175[degrees]C were 76% and 51% respectively. Mid-range low temperature properties were observed. The margin of safety, however, with regard to these properties was still rather small. Higher viscosity polymers appeared to improve aged elongation and compression set. This, however, gives rise to processability concerns particularly in formulations with low plasticizer content. The compound viscosity in Recipe #38 is 89 Mooney units. This approaches the upper limit for many commercial scale processes.

A designed experiment was used to determine if additive changes could improve the balance of properties while maintaining processability. From the polymer effects study, it was found that a higher viscosity polymer system with medium ethylene content had a good balance of properties and good extrusion characteristics. The base polymer used in the designed experiment was held constant at 50% EPDM A and 50% EPDM C. This blend had a composite 62% ethylene concentration, 1.2% ENB and was 52 Mooney, ML 1 + 4 (125[degrees]C).

Three variations were made to the base formulation. The plasticizer types, antioxidant system and coagent type were varied in a 2 x 3 full factorial experimental array. The test formulations used and data obtained are given in table 4.

[TABULAR DATA OMITTED]

Liquid polybutene was evaluated as a substitute plasticizer for paraffinic oil. Since many of the previous studies found that plasticizer content must be minimized to improve aging resistance, it was reasoned that thermal or oxidative instability of the paraffinic oil used contributed to the problem. Plasticizer volatility problems will also contribute to aged elongation and compression set losses.

Polybutene was chosen as a potential alternative plasticizer for several reasons. First, because it contains no aromatic or allylic carbons, it should be more difficult to form free radicals associated with oxidation. Second, it is not expected to actively crosslink with the EPDM polymer. Lastly, polybutene is compatible with EPDM and is relatively nonvolatile. The polybutene used had an average weight of 2060. A high molecular weight plasticizer was chosen to minimize volatility.

A hindered phenol stabilizer, octadecyl 3,5-di-t-butyl-4-hydroxyhydrocinnamate, was added to the current antioxidant package to determine its effect. Hindered phenol stabilizers are generally not used in formulations containing carbon black. However, because it is a stabilizer used in many polyolefin processes, its effect in this formulation was studied. It was suspected that stabilizer chosen by the polymer manufacturer may play a role in stabilization at these temperatures.

Zinc diacrylate (ZDA) and trimethylol propane trimethacrylate (TMPTM) were compared as coagents. ZDA is less reactive than TMPTM. This coagent was investigated to determine if processing safety benefits, aging resistance and/or compression set could be improved.

Results and discussion

Base polymer effects

Several relationships were observed between base polymer ethylene content and stress strain properties. As would be expected, tear resistance, modulus and hardness increase with increasing ethylene content. Weaker relationships with ultimate elongation and tensile strengths were also observed. Elongations decreased with increasing ethylene content. Tensile strength increases with increasing ethylene content. No strong base polymer molecular weight effect on original stress strain was observed.

As indicated in table 3, tensile sheets for each compound were aged for one and two weeks at 150[degrees]C and one week at 175[degrees]C. Compression set measurements were made after aging 70 hours at 150[degrees]C and 70 hours at 175[degrees]C. In these tests, polymer molecular weight and ethylene content effects were observed (figures 2, 3 and 4). Higher molecular weight polymers tended to have better elongation retention properties while the higher ethylene polymers tended to have better aged hardness retention. Interestingly, these effects appeared to be independent of each other.

[CHART OMITTED]

One could argue that the ethylene effect on aged hardness retention may be partially attributed to limitations of the hardness test itself. Most of the formulations were in the 85 to 90 range after aging. The Shore A hardness test tends to lose discernability as hardness approaches 90. Since original hardness values also approach the 85 to 90 range, it will be necessary to validate this effect by studying lower hardness formulations. This will be addressed in future work.

Molecular weight was found to have an important effect on compression set resistance. This effect indicates that lower compression set is obtained with higher molecular weight polymers. This effect is attributed in part to the increased elasticity observed with increased molecular weight (ref. 7). However, since this phenomenon deals with a crosslinked system, there may also be a beneficial effect of favorable unsaturation distribution.

Because the number of unsaturation sites in EPDM is low, polymer molecular weight has an effect on the number of active crosslinking sites per chain.

Ease of free radical formation associated with peroxide crosslinking will follow the following hierarchy:

Benzyl > allyl > 3[degrees] > 2[degrees] > 1[degrees] > vinyl, [CH.sub.3].

For this reason, peroxides will preferentially crosslink at the allylic hydrogen sites associated with the ENB monomer. The distribution of ENB will therefore affect the distribution of crosslinks. If one considers that ENB incorporation to the polymer chain is random, a model can be used to illustrate that higher molecular weight polymers have more unsaturation sites per chain at constant total unsaturation. Take for example the following polymer chains:

A: [[EP(ENB)EEPPEPEPEEEPEEPEPPEE].sub.500]

B: [[EP(ENB)EEPPEPEPEEEPEEPEPPEE].sub.1,000]

Both polymer A and B are identical in terms of relative monomer composition. Polymer B, however, having higher molecular weight, has twice the number of ENB monomers per chain. Since good compression set resistance is generally attributed to high number of crosslinks per chain, Polymer B would be expected to have lower compression set.

Low temperature properties were also studied using temperature retraction. As would be expected, low temperature properties deteriorate when ethylene content increases. The temperatures at 10% recovery are illustrated in figure 5.

[CHART OMITTED]

Extrusion studies were conducted on a 13 mm extruder at constant screw speed and a temperature of 104[degrees]C. All formulations extruded very well as evidenced by Garvey die ratings. In this configuration, extrusion rates obtained are a function of compound viscosity. Low viscosity formulations had faster rates. Rate effects will vary with differing extruder configurations. Surface appearance, feed characteristics and dimensional consistency were obtained with these formulations over a wide viscosity range.

The best balance of properties in the polymer effects study clearly favors the use of higher molecular weight polymers. Better elongation retention and compression set are the main benefits. Higher ethylene polymers may have some benefits in aged hardness retention, tear resistance, tensile strength improvements and other benefits. However, low temperature properties will possibly suffer if ethylene levels are too high (particularly if plasticizer content is minimized).

ENB content should not be overlooked as a critical variable. Although low unsaturation polymers were used predominantly in this study, three formulations (#32, #39 and #43) had 2.4% or higher unsaturation. All had poor aged elongation properties. The other formulations where ENB levels varied from between 1% and 2% had much better properties.

Compound optimization studies

As shown by the extrusion data in table 4, all experimental in the plasticizer design processed well. There was no evidence of plasticizer bloom in compounds containing polybutene. Unfortunately, compound viscosities were rather high, as would be expected due to the choice of base polymers. A summary of the statistical treatment of data is given in table 5.

[TABULAR DATA OMITTED]

In general, the best aging properties were obtained using a combination of the polybutene plasticizer and the TMPTM coagent. Compounds containing polybutene averaged 10% lower compression set and 11% higher elongation retention than formulations containing paraffinic oil.

Compounds containing ZDA had similar aged elongation but much poorer compression set than compounds containing TMPTM. However, ZDA had a beneficial effect on original elongation and tear resistance.

The results indicate that the presence of the hindered phenol stabilizer significantly interferes with peroxide cure efficiency. Increased original elongation and lower modulus and poorer aging resulted from addition of this AO.

Polybutene aging performance at 175[degrees]C was markedly better than that of paraffinic oil. At 150[degrees]C, however, no significant difference was observed. This may be an indication that the thermal/oxidative stability of the paraffinic oil deteriorates rapidly between 150[degrees]C and 175[degrees]C. Volatility probably also increases significantly over this range.

Since the aging results of compounds containing polybutene are similar to those obtained in the plasticizer free formulations reported by Dunn (ref. 1), one can conclude that the thermal/oxidative stability and volatility of the polybutene selected approaches that of the EPDM base polymer. Further study of the effects of polybutene molecular weight and addition level is warranted.

Lower original elongation and beneficial compression set without hardness increase suggest that the polybutene plasticizer interferes with peroxide activity to a much lesser extent than paraffinic oil. Again, further study of this effect is necessary before firm conclusions can be made.

Since beneficial effects are seen by reducing the content of aromatic carbons associated with paraffinic oil plasticizers, there may also be some benefit from reducing other sources of reactive carbons. Clearly, the surface chemistry of carbon black is a logical target. Beneficial effects of silane treated talc as a partial replacement for carbon black were cited in several reports (refs. 1, 2, 5 and 6). Carbon black elimination might serve to improve aging resistance. This speculation is consistent to results obtained from the EPDM polymer studies. The allylic hydrogens associated with ENB are more susceptible to oxidation than the polymer backbone. Minimization of propylene monomer content will reduce the concentration of tertiary hydrogens which are more susceptible to oxidation than secondary hydrogen. Of course, there is a balance as peroxide vulcanization relies on the formation of free radicals to form bonds between chains. This is perhaps why some ENB content in the polymer backbone appears to be beneficial.

Summary

Through this investigation, 175[degrees]C aging characteristics of EPDM formulations with practical processing characteristics were improved beyond those observed in previous work (ref. 2). Targets for aged stress strain properties and compression set were achieved with a greater margin of safety.

* Base EPDM polymers with higher molecular weight had better aging characteristics.

* A polybutene plasticizer provided better aging resistance than did paraffinic oil. Further investigation of the effects of polybutene molecular weight and addition level is warranted.

* Base EPDM polymer unsaturation should be 2% or lower for good aged elongation retention.

* Base EPDM polymer ethylene content should be moderately high to effect good tear resistance and hardness retention without substantial sacrifice to low temperature properties.

* Further investigation to eliminate sources of benzylic and allylic carbons within the formulations may result in improved aging performance. A possible source is carbon black.

References

(1.)Dunn, J.R., "Compounding ethylene-propylene elastomers for high temperature aging resistance," International Rubber Conference, Harrogate, England (June, 1987).

(2.)Dunn, J.R., Keller, D.R. and Paterson, J., "EPDM compounding for high temperature aging requirements in automotive hose," ACS Rubber Division, Cleveland, OH (October, 1987).

(3.)Ashworth, B.T. and Hill, P., "Some observations on unusual antioxidant effects," International Rubber Conference, Nurembourg, W. Germany (September, 1980).

(4.)Scott, G. and Tavakoli, S.M., "Polymer degradation and stability," 19 (1987) 29-50.

(5.)Keller, R.C., SAE technical paper #870190 "Meeting coolant hose compound requirements of the future," (February, 1987).

(6.)Hudson, J.C., "Automotive under the hood EPDM compounds," Southern Rubber Group Meeting, Ashville, NC (February, 1990).

(7.)Gallagher, M.T., "EPDM microstructure effects in polypropylene," Society of Plastics Engineers Conference, Montreal, Quebec (May, 1991).

(8.)Morrison, R.T. and Boyd, R.N., "Organic Chemistry," 3rd Edition, Allyn and Bacon (January, 1975).