Standardization of EPDM characterization test for QC and specification purposes.

In the range of commercial elastomers the importance of ethylene-propylene-rubber (EPM) and ethylene-propylene-diene rubber (EPDM) has increased significantly in the 30 years since these synthetic rubbers were introduced on the market. EP(D)M presently accounts for roughly 10% of total worldwide synthetic rubber consumption (ref. 1). Its main use is in high-quality applications, particularly automotive hoses and seals; in building and construction in profiles, roofing foil and seals; in cable and wire production in cable insulation and jacketing; and in appliances in a wide variety of - usually molded - articles. In addition, considerable amounts of EPDM rubber are used in blends with thermoplastics, e.g. as impact-modifiers or as a base material for thermoplastic elastomers.

Even though EPDM rubbers have been commercially available for quite some time, the technology concerning these products - both their production and their applications is still very much under development. All EP(D)M manufacturing processes are highly proprietary and differ greatly between various suppliers (ref. 2).

As a result, the various EPDM suppliers' test methods for characterizing the specific polymers have also evolved in different directions. Efforts have been made to standardize evaluation procedures for EPDM in collaboration with standardization institutes such as ASTM (ref. 3), DIN (ref. 4) and ISO (ref. 5). But these standards are restricted to Mooney measurements and vulcanization characteristics or vulcanized properties based on one or more standard compound recipes. The precision of the Mooney measurement has been improved greatly over the last few years throughout the whole rubber industry (refs. 6-8). Yet the Mooney viscosity test is a rather restricted technique, which can hardly be considered capable of predicting the processability of EPDM for the wide range of processing techniques employed in the rubber industry. There is generally felt to be a need for another or several other characterization technique(s) to gain better control of the variation in the processability of polymers with what are otherwise the same Mooney viscosities (ref. 9). It is generally recognized that, as regards measurements of vulcanization characteristics using curemeters and/or vulcanized properties of plaques which are usually press-cured, the precision of the tests is poor, largely as a result of the variability introduced by the necessary compounding and mixing steps. The use of these properties for specification purposes therefore has little or no meaning, while the specification ranges are almost completely determined by the (lack of) precision of the test method rather than by the actual variation in the product.

The properties of EP(D)M rubbers are dependent on a number of structural parameters of the copolymer chains:

* the relative content of comonomer units and the way these are distributed in the chain;

* the type and amount of unsaturation introduced by the "third monomer" needed for sulfur vulcanization;

* the average molecular weight and molecular weight distribution;

* long-chain branching, all of which can be regulated via the operating conditions during polymerization and via the chemical composition of the catalyst. It is as a result of the differences between the various processes that all EP(D)M polymers from different suppliers show subtle differences. And present state-of-the-art polymer characterization, however sophisticated, in many cases fails to detect these differences to a sufficient extent.

In 1987 the Wirtschaftsverband der deutschen Kautschuk-industrie (W.d.K.) embarked on an attempt to develop standard formats for quality specifications for the various raw materials used for rubber compounding in Germany. The producers of these raw materials were asked to cooperate. The formats included characterizations for specification purposes, for certification and for identification of the raw materials.

When the EPDM producers had been approached, the great variety of test methods employed by the different producers hampered the process of reaching common specification formats for EPDM. The joint European EPDM suppliers, coordinated by the IISRP European section, recognized the need for standardization of EPDM characterization tests for quality-control and specification purposes. To this end a working group was established, which over a period of several years developed a set of "best practice" test methods for EPDM. This was done either on the basis of standardized test methods or on the basis of methods generally accessible in the literature or by selecting a method which enabled the various participants to generate comparable data on common samples with good repeatability.

The results of these activities are described in this article. Any decision to implement such test methods as well as the timing thereof are left to the discretion of the individual suppliers and consumers.

Description of the molecular composition of EPDM

EPM and EPDM are copolymers of ethylene, propylene and, in the case of EPDM, of a third monomer, a chemical group offering a carbon-carbon double bond for sulfur vulcanization. Commonly used third monomers today are ethylidene norbornene, dicyclopentadiene and 1,4-hexadiene. These third monomers are used alone or, in some cases, in combinations.

The ethylene content of EP(D)M ranges between about 45 wt% and about 75 wt% between different grades. It is an important basic property related to the green strength of both the pure polymer and the compound, the extrusion characteristics of the compound and a variety of properties of the cured material.

The contents of the third monomers usually range between 0 wt% and roughly 12 wt%, depending to a certain extent on the type of third monomer. This basic property obviously has a major influence on the vulcanization behaviour of a particular EPDM grade.

Depending on the emphasis put on the respective characteristics of EP(D)M, there are various ways used today to describe the composition of EP(D)M:

a. Ethylene content (wt%) + propylene content (wt%) = 100%; third monomer content is taken as additional to the 100 wt%.

b. Ethylene content (mol%) + propylene content (mol%) = 100%; third monomer content is taken as additional to the 100 mol%.

c. Ethylene content (wt%) + propylene content (wt%) + third monomer content (wt%) = 100%.

d. Ethylene content (mol%) + propylene content (mol%) + third monomer content (mol%) = 100%.

e. Various other, less common combinations.

There are different reasons for using each of these modes. For example, the absolute ethylene content expressed in mol% is the governing factor for the above mentioned green-strength properties of the pure polymer and compounds thereof. However, in the case of b, the ethylene content has to be corrected for the amount of third monomer (in mol%) in the molecule to achieve the absolute ethylene content of the polymer. This includes a calculation based on the different molecular weights of the third monomers used.

The different methods used to describe the composition of EP(D)M are generally felt to be confusing, in particular because it is not always clear which description method is being used.

For this reason the IISRP working group recommends that the composition of EP(D)M be defined as follows:

Ethylene content (wt%) + propylene content (wt%) + unsaturation content (wt%) = 100%

The choice of the particular method used to describe the molecular composition is left to the discretion of the supplier or consumer.

Experiments

Materials

Standard samples for various comparative purposes throughout this study were selected mainly from commercially available polymers:

Sample 1: Buna AP 248 Huls AG

Sample 2: Buna AP 331 Huls AG

Sample 3: Buna BA 648 Huls AG

Sample 4: Keltan 509x100 DSM Elastomers Europe

Standard calibrants developed in the course of this study:

For ethylene content:

Standard 7: Vistalon 404 Exxon Chemical Europe

Standard 8: Buna AP 301 Huls AG

Standard 9: Dutral CO 03X Enichem Elastomeri Spa

Standard 10: Vistalon 805 Exxon Chemical Europe

For unsaturation content. ENB and DCPD:

Standard 1: Vistalon 8175 Exxon Chemical Europe

Standard 2: - Exxon Chemical Europe

Standard 3: Vistalon 5600 Exxon Chemical Europe

Standard 4: Vistalon 8504 Exxon Chemical Europe

Standard 5: Keltan 52() DSM Elastomers Europe

Standardization of laboratory test methods

The approach taken to standardize EP(D)M test methods between the participants of this study generally conformed to ISO Standard 5725 (ref. 10) The definitions of precision. repeatability and reproducibility are given in this standard and will be used in this article accordingly The exercise is therefore to be considered as a standardization exercise and not as an exhaustive attempt to quantity exactly the precision of the test methods. for which many more participating laboratories and tests would have been required As stated in ISO 5725, this next step can be carried out only once a test method has reached a general level of acceptance. The results of all efforts are therefore restricted as regards their level of accuracy.

The experiments in this article should in general be considered as an attempt to determine the repeatability of the various test methods. carried out under repeatability conditions in a particular participating laboratory by the same operator using the same equipment within a short interval of time. For practical reasons the repeatability data are presented here as the (repeatability) standard deviation of the test results within the same laboratory and in terms of the (repeatability) coefficient of variation In all cases tests were carried out on identical materials, as described in the preceding section. The results are not presented as reproducibility, to include variability between laboratories: including differences arising from the use of various operators, different equipment. deterrences in average test results, etc. By showing the actual average test values and repeatability data per laboratory, we ask the reader to form his/her own opinion on the inter-laboratory precision of the test methods.

It should also be borne in mind that the exchange of test methods between the different participants meant that in many cases some laboratories were carrying out such tests for the first time. The obvious lack of experience resulted in somewhat poorer repeatability than would have been the case had these laboratories been more experienced. Therefore, it is to be expected that, once such test methods become generally accepted, the repeatability will further increase. Similarly, all experimental data were included in the calculation of the repeatability, and no corrections were made for so-called outliers or stragglers. No particular attention was paid to the possibility of bias of the test methods, other than where explicitly mentioned (oil content) or where an effort was made to abolish a known bias due to a historic selection of biased standard calibrants (ethylene determination). In general it was assumed that the test results are normally more or less distributed around an average and that therefore Gaussian statistics apply.

Ethylene ([C.sub.2]-) content of EP(D)M

The method most commonly used to determine the molecular composition of EP(D)M is infrared spectroscopy (ref. 11). The ASTM D3900 test method was established on this basis in 1980 by a joint project in the North American EP(D)M production industry (ref. 12). It makes use of the relative absorbance of characteristic bands within the infrared spectrum of sequential [-CH.sub.2]-units versus [-CH.sub.3]-units. Depending on the relative amount of ethylene vs. propylene in EP(D)M, it is common practice to use different combinations of absorbance peaks. The particular choice of peaks, to a certain extent determined by historic choices and experience within the different laboratories, caused differences between the various EPDM suppliers. In addition, other variants of this technique were used within the European EP(D)M production industry for various reasons.

As infrared spectroscopy is a relative technique, it needs calibration using absolute data. A set of six EPM standard calibrants were developed for the ASTM D3900 standard, as issued in 1980, of which the ethylene contents had been determined on the basis of NMR measurements of [sup.14.C]-tagged pilot-plant polymers. It is generally accepted that NMR, and particularly [sup.13.C]-NMR, is an absolute technique to measure the ethylene content of EP(D)M. The ASTM standard calibrants were not generally accepted, and in fact most producers used their own sets of calibration standards from various sources. It was therefore felt that agreement needed to be reached on a new set of common standard calibrants, or rather a commonly applicable up-to-date NMR technique, so that necessary additional calibration standards could be prepared without experimental NMR problems arising.

Both aspects of the ethylene determination of EP(D)M are covered below.

Infrared determination of the ethylene content of EP(D)M

In an initial attempt five participating laboratories measured the propylene ([C.sub.3]-) content of three selected EPDM samples to compare their own in-house test methods for repeatability and average values obtained. The results are given in table 2. It is quite obvious that rather large differences were obtained by the various laboratories in average numbers as well as in the repeatability of the measurement as expressed in the coefficient of variation. Part of the reason for the differences in average numbers was the lack of a unified set of calibration standards.

Table 2 -- propylene content of three selected EDPM samples, measured by own in-house method

                 Sample 1            Sample 2          Sample 3
               (x)    (s)  (v)   (x)   (s)   (v)    (x)   (s)    (v)
Lab 1          29.9  .08  .27   49.4   .12   .23   23.7   .11   .46
Lab 2          31.0  .21  .68   49.8   .25   .50   24.7   .18   .73
Lab 3 (using                    52.8   .05   .09   19.3   .20  1.04
ASTM D3900)
Lab 4          28.6  .20  .70   46.1   .30   .65   25.5   .10   .39
Lab 5          30.0  .80  2.7   50.6  1.50  3.0    23.1  1.15  4.9

(x) = average (mass %); (s) = standard deviation (mass %) (v) = coefficient of variation (%)

These results were considered quite unsatisfactory. An attempt to improve the situation by having all the laboratories adopt one of the best-performing in-house methods did not result in a significant improvement. The fact that one laboratory, using ASTM standard D3900, achieved a relatively good repeatability prompted all the participating laboratories to repeat the exercise using ASTM D3900 and the six calibration standards as the actual samples to be measured and used for the calibration at the same time. The results are shown in table 3. Quite clearly, all the participating laboratories achieved closer results, not only for the averages but also in general as regards better repeatability, as expressed by the coefficient of variation. The closeness of the averages is obvious, of course, because all the participating laboratories used the same samples for the measurements as well as for calibration purposes.

Table 3 - ethylene content of the six calibration samples, measured by ASTM standard D3900

           Standard 1                        Standard 2
           x       s     v                        s      v
Assigned [C.sub.2] content (mass %)
          37                             48
ASTM D3900 method A:

Lab 1    36.67   .113   .31                      .209   .43
n=8
Lab 2    36.68   .120   .33                      .240   .49
n=8
Lab 3    36.63   .063   .17                      .114   .23
n=8
Lab 4    36.79   .049   .13                      .038   .08
n=8
Lab 5    37.0    .053   .14                      .307   .65
n=8
Lab 6    36.7    .097   .26                      .118   .24
n=10
ASTM D3900 method B:
Lab 2
n=8
Lab 3
n=8
Lab 4
n=8
Lab 5
n=8
Lab 6
n=10

           Standard 3                        Standard 4
           x       s     v               x        s      v
Assigned [C.sub.2] content (mass %)
          55                             63
ASTM D3900 method A:

Lab 1    55.37   .180   .33              62.81   .214   .34
n=8
Lab 2    54.59   .329   .60              65.72   .403   .16
n=8
Lab 3    55.29   .089   .16              62.55   .213   .34
n=8
Lab 4    54.94   .062   .11              62.86   .021   .03
n=8
Lab 5    55.5    .181   .33              63.0    .717  1.20
n=8
Lab 6    55.1    .206   .37              62.7    .422   .67
n=10
ASTM D3900 method B:
Lab 2                                    65.37   .672  1.03
n=8
Lab 3
n=8
Lab 4
n=8
Lab 5                                    66.6    .818  1.23
n=8
Lab 6                                    63.4    .512   .81
n=10

           Standard 5                        Standard 6
           x       s     v               x        s
Assigned [C.sub.2] content (mass %)
          69                             79
ASTM D3900 method A:

Lab 1    68.82   .349   .51              78.40   .223
n=8
Lab 2    68.08   .404   .59              78.04   .357
n=8
Lab 3    55.29   .089   .16
n=8
Lab 4    54.94   .062   .11
n=8
Lab 5    70.2    .511   .73              80.0    .714
n=8
Lab 6    55.1    .206   .37
n=10
ASTM D3900 method B:
Lab 2    70.08   .861  1.23              79.67   .291
n=8
Lab 3    69.00   .176   .26              79.0    .088
n=8
Lab 4    69.00   .090   .13              79.0    .109
n=8
Lab 5    69.7   1.02   1.46              79.1    .052
n=8
Lab 6    68.5    .931  1.36              79.2    .327
n=10

The conclusion can be drawn that all the participating laboratories managed to generate more or less the same average numbers with a comparable, acceptable repeatability on the basis of the description of the standard ASTM test method and the corresponding standard calibrants. The coefficient of variation is nearly always lower than 1%, which is an extremely good result compared with typical repeatability figures obtained on, for example, mechanical properties testing in the rubber industry. Based on these encouraging results it was agreed that this method should be recommended as the preferred method for the determination of the ethylene content of EP(D)M polymers.

ASTM D3900 Standard Test Method makes a distinction between two different methods, making use of pressed film samples, using regular infrared wavelengths:

Method A - For EP(D)M between 35 and 65 mass% ethylene:

Method B - For EP(D)M between 60 and 85 mass% ethylene;

where there is an overlap between the two in the range of 60 - 65 mass% ethylene. The difference is based on another selection of infrared peak ratios:

Method A - 1,160/720 [cm.sup.-1] peak ratio;

Method B - 1,378/720 [cm.sup.-1] peak ratio;

because for EP(D)M polymers with higher mass percentages of ethylene the 1,160 [cm.sup.-1] peak tends to become too small to ensure sufficient accuracy. This was obviously based on the experiences with the quality of infrared equipment at the time that the ASTM Standard was drawn up, i.e. before 1980.

The results of table 3 tend to indicate that with present day infrared equipment with method A acceptable averages and repeatability can be achieved, not significantly worse than with method B, even for the high ethylene standards 5 and 6. If one method could be used rather than two, avoiding a separate calibration, it would greatly simplify matters and possibly add to the overall accuracy of the determination. This was looked at more closely by four laboratories measuring ASTM calibration samples 4, 5 and 6 once more, according to both methods A and B, as shown in table 4.

Table 4 -- comparison of ASTM standard D3900 method A and method B for three standard calibrants with a high ethylene content

              Standard 4               Standard 5
Lab 1        x       s      v         x       s    v
Method A    63.0    .40   .63        69.2    .28   .41
Method B    63.4    .51   .80        68.5    .93  1.36
Lab 2
Method A    66.3    .21   .32        72.3    .35   .48
Method B    Difficulties with film   72.0    .44   .61
Lab 3
Method A    62.8    .23   .37        69.5    .29   .42
Method B                             69.0    .28   .42
Lab 4
Method A    63.0    .72  1.20        70.2    .51   .73
Method B    66.6    .82  1.23        69.7   1.02  1.46

                Standard 6
Lab 1          x       s    v
Method A      78.6    .28   .36
Method B      79.2    .33   .42
Lab 2
Method A      81.9    .16   .20
Method B      82.0    .16   .20
Lab 3
Method A      79.0    .18   .23
Method B      79.0    .22   .28
Lab 4
Method A      80.0    .71   .89
Method B      79.1    .05   .06

Again, there is no particular reason to prefer method B to method A. This would suggest the adoption of method A as the only single method. However, when applied to a series of commercial polymers, differences were observed for which no explanation could be given other than interferences due to additives and stabilizers that influence the weak 8.65 m peak at these high ethylene contents, as already indicated in the ASTM Standard D3900. For this reason the members did not reach agreement on this issue. But there are no objections to the use of method A only, wherever it is possible, provided it is indicated in the description of the test method.

The IISRP working group decided to recommend ASTM Standard Test Method D3900 for the determination of the mass % ethylene of EP(D)M. Either method A and/or method B may be used, depending on personal preference, provided that the method is indicated with the test results.

Calibration standards for ASTM D3900

As indicated before, the six standard ASTM D3900 calibrants date back to before 1980. Practical experience with these standard calibrants had led to suspicion concerning the ethylene contents assigned to these standards. There were indications that, in comparison with mass balance measurements in the EPDM production processes, these standards were sometimes significantly off-set. This is one of the reasons why several producers developed their own in-house calibrants, considered to be more realistic (see previous section).

As a major step in reaching agreement on a standard set of calibrants among the participants of this working group, a detailed [sup.13.C]-NMR procedure was developed to analyze ethylene-propylene rubbers. Special attention was paid to instrumental and spectral parameters and calculation methods. This work was submitted for publication in the scientific literature (ref. 13). By applying this procedure, it is possible to obtain consistent and reproducible molecular composition results for EPM rubbers, independent of the measuring laboratory and the calculation method. The procedure was tested in eight industrial laboratories.

In a joint project between this working group and a consortium of North American EPDM producers, represented by ASTM Committee D11.11, the original six 1EPM standard calibrants of ASTM Standard Test Method D3900 were recalibrated, and another four were added to form a total of ten calibrants. These ten new standard calibrants were adopted by ASTM and published in the July 1994 issue of ASTM Standard D3900 (ref.14).

Consequently, the re-established mass percent values for the original six standard calibrants differ to some extent from the values assigned in 1971/1980, as depicted in table 5 and figure 1.

[Figure 1 ILLUSTRATION OMITTED]

Table 5 -- standard calibrants for ASTM D3900 for the determination of mass % ethylene in EP(D)M, rubbers, old and newly assigned values.

Standard number    Mass percent ethylene
                   Old               New
1                  37               40.1
2                  48               52.4
3                  55               58.6
4                  63               66.8
5                  69               70.8
6                  79               78.6
7                  --               44.8
8                  --               52.6
9                  --               69.5
10                 --               77.5

The use of these new mass percent ethylene calibrants will result in ethylene values being calculated that differ from those previously expected, giving differences up to 4.5 mass% in certain cases.

Table 1 - participating companies

Bayer/Polysar
DSM Elastomers
   Europe BV
Du Pont, German office
Enichem Elastomeri
Exxon Chemical
   Europe
Huls AG

References

[1.] Worldwide rubber statistics, International Institute of Synthetic Rubber Producers Inc., Houston, TX.

[2.] J.W.M. Noordermeer, Kirk-Othtner Encyclopedia of Chemical Technology -- Fourth Ed., 8, 978 (1993), John Wiley & Sons, Inc.

[3.] ASTM D3568-90. Standard Test Methods for Rubber - Evaluation of EPDM (ethylene propylene diene terpolymers) including mixtures with oil.

[4.] Prufung von Kautschuk und Elastomeren, DIN 53 670 Teil 10: Prufung von Kautschak in Standard Testmischungen; ethylen-propylen-dien kautschak EPDM, 1983 issue.

[5.] International Standard, ISO 4097 rubber, ethylene-propylene-diene (EPDM), non-oil extended raw general-purpose rubber - evaluation procedures, 1991 issue.

[6.] J. Markert, Gummi Asbest Kunststoffe, 9, 568 (1976).

[7.] R. Koopmann, Kautschak + Gummi, Kunststoffe, 38, 281 (1985).

[8.] H. Kramer, Kautschak + Gummi, Kunststoffe, 43, 912 (1990); Rubber World, 204, 35 (1991).

[9.] W. Breemhaar, R. Koopmann, J. Markert and J. Noordermeer, Kautschak Gurnmi Kunststoffe, 46, 957 (1993).

[10.] International Standard, ISO 5725, Precision of test methods - determination of repeatability and reproducibility for a standard test method by inter-laboratory tests.

[11.] I.J. Gardner, C. Cozewith and G. Verstrate, Rubber Chemistry and Technology, 44, 1015 (1971).

[12.] Standard Test Methods for Rubber, ASTM D3900, determination of ethylene units in EPM and EPDM, 1980 issue.

[13.] S. DiMartino and M. Kelchtermans, Determination of the composition of ethylene propylene-rubbers using 13C-NMR spectroscopy, J. Appl. Poly. Sci in print.

[14.] Standard Test Methods for Rubber, ASTM D3900, determination of ethylene units in EPM and EPDM, 1994 issue.

Acknowledgements

"The recycle of plastics and rubber - a contrast" is based on a paper given at the October, 1996 Rubber Division meeting."

"New technology for the devulcanization of sulfur-cured scrap elastomers" is based on a paper given at the October, 1996 Rubber Division meeting.

"New solid state shear extrusion pulverization process for used tire rubber recovery" is based on a paper given at the October, 1996 Rubber Division meeting.

This is the first installment of a two-part series. Dr Noordeermeer authored this on behalf of the Inaternational Institute of Synthetic Rubber Producers. The objective is to achieve standardization among EPDM producers of testing procedures for the basic properties. This article was originally published in Kautschuk + Gummi, Kunstsoffe. The second part will run in July.