Printer Friendly

Improved peroxide curing efficiency of EPDM with VNB as third monomer in an engine mount formulation--part 2.


Compound--network density--properties correlations The set of peroxide-cured high VNB-EPDMs in combination with the ENB-EPDM references prepared in this study provide a comprehensive collection of experimental data for a more in-depth investigation of compound formulation [left and right arrow] network structure [right and arrow] vulcanizate properties correlations. Such an investigation was shown to be very fruitful in our first study on peroxide-cured EPDM in a window gasket formulation (ref. 18). In the previous section, correlations between compound formulations (EPDM type and curative package) and vulcanizate properties have been established in a direct approach. The intermediate step, i.e., the network formation, is only included intuitively, but not explicitly. In a more scientific approach, it is first considered that the compound formulation results in a certain crosslinked network with crosslink density being one of the main characteristics, which in its turn results in a set of vulcanizate properties. In addition, it will be shown that hot air aging results in additional crosslinking, which can be described by the same network structure [left and right arrow] vulcanizate properties correlations. Such step-by-step correlations do not only provide insight into peroxide crosslinking in a merely scientific way, but can also be used to guide future product and application developments.

Compound formulation [right arrow] crosslink density

The correlations between compound formulation and cross link density will be discussed in detail first. The main parameters determining the crosslink density in these peroxide-cured compounds are the diene type (VNB or ENB), the diene content and the level of curatives, as already demonstrated by the rheometry data (see Part 1, table 3 and figure 3 (RW+)). In rubber technology, the difference in rheometer torque (MH-ML) is traditionally used as a measure of crosslink density. A real time NMR relaxation time study on crosslinking of EPDM has shown that (MH-ML) is a quantitative measure of crosslink density (ref. 23). Figure 6 shows that (MH-ML) has curved correlations with the Px (peroxide) level for each EPDM type. The peroxide curing efficiencies increased in the sequence: ENB-EPDM < 1% VNB-EPDM < 2% VNB-EPDM < 3.6% VNB-EPDM, which confirms first of all that VNB-EPDM has a high reactivity for peroxide cure, and secondly that increasing diene monomer content results in enhanced crosslinking. Extrapolation of the curves to zero Px level yields (MH-ML) values of zero, which simply confirms that there is no crosslinking in the absence of peroxide.


Figure 7 demonstrates that a plot of (MH-ML) against the content of the main diene monomer gives good linear correlations for the various VNB-EPDMs. Upon increasing the Px level, the (MH-ML) against VNB content correlations are not only shifted upward, but also become more steep. To understand these phenomena, it is reminded that peroxide crosslinking of EP(D)M proceeds via two reaction pathways, i.e. combination and addition (Part 1, scheme 1), and that the total peroxide crosslink density is the sum of the contributions of both (refs. 6 and 7). The extent of both combination and addition reactions will depend on the peroxide level, whereas only the extent of the addition reaction is affected by the diene type and content. Extrapolation of (MH-ML) to zero VNB content yields the crosslink density of EPM without diene, which of course deviates from zero. These extrapolated values represent the contributions of the combination reaction to the total peroxide crosslinking and increase with increasing Px level. Incorporation of diene monomers will result in additional crosslinking via the addition reaction, resulting in a higher (MH-ML). The increase of the slopes of (MH-ML) against the diene content indicates that increasing the peroxide level results in enhanced addition reactions. The (MH-ML) values for ENB-EPDM do not fall on the linear correlations for the three VNB-EPDMs (points at 4.5% diene in figure 7). The (MH-ML) values for ENB-EPDM are much lower, which shows that the peroxide curing efficiency of ENB-EPDM is smaller than that of the VNB-EPDMs. From previous studies it is known that the peroxide curing efficiency of VNB-EPDM is about four times higher than that of ENB-EPDM. The arrows in figure 7 indicate a shift by a factor of four of the diene content of the ENB-EPDMs, i.e., 4.5 wt. % ENB agrees with ~1.1 wt. % VNB. As a result of this horizontal shift, the data for ENB-EPDM have a reasonable fit with the VNB-EPDM correlations. The lack of fit is explained by the fact that the ENB-EPDM has a rather low number-averaged molecular weight (table 1) compared to the 1% and 2% VNB-EPDMs, which will result in less effective network formation, and thus a somewhat lower (MH-ML).




It is concluded that the crosslink density of peroxide-cured EPDM is determined by both the Px level and the EPDM diene type and content. To allow a straightforward comparison between VNB-EPDM and ENB-EPDM, the ENB content is converted into so-called VNB equivalents by dividing by the factor 4. The four EPDMs studied contain either VNB as main diene monomer with a small amount of ENB or the other way around, which are combined yielding the total VNB equivalent. In a first approximation, the product of the Px level and the total VNB equivalent will be used as the single parameter determining the susceptibility of the EPDM compounds for peroxide cure. In figure 8, (MH-ML) is plotted versus this product [Px * (total VNB equivalent)] and a fair correlation is obtained. Considering the simplicity of this approach and the fact that the polymers not only differ in diene type and content, but also in molecular weight (distribution) and degree of LCB (table 1), the correlation is considered satisfactory. Figure 9 shows a contour plot of (MH-ML) versus Px level and total VNB equivalent content ([R.sup.2] = 0.92). The maximum difference between the calculated and the experimental (MH-ML) is less than 0.5 Nm, which shows that the prediction is similar to the experimental error for (MH-ML) on an ODR2000 rheometer.

Crosslink density [right arrow] properties

In the previous section, correlations between compound formulation and crosslink density have been established. Now, the correlations between the crosslink density and the properties of the peroxide-cured EPDM compounds will be discussed in more detail. Figures 10a-d show plots of some selected properties (M300%, TS, EB and CS125[degrees]C) of the peroxidecured compounds against rheometer (MH-ML). The general conclusion is that for all the properties, a single correlation exists between the particular property and (MH-ML). This shows that the crosslink density is indeed the main parameter governing the properties of these peroxide-cured compounds, and that the details of the network structure are of minor importance. The established correlations can be used as master curves to predict properties for a given crosslink density or, in a reversed approach, to estimate the crosslink density required for a desired combination of properties. Obviously, these specific master curves are only valid for the current engine mount compounds, although the general principle also holds true for other compound formulations, as was already shown in our first study on window gaskets (ref. 18).



M300% of the peroxide-cured compounds increased with (MH-ML) and shows a linear correlation (figure 10a), which is in agreement with the statistical network theory of rubber elasticity (ref. 24). As expected, HD (hardness) being a sort of surface modulus also increased with (MH-ML) in a linear fashion. TS showed one single master curve versus (MH-ML) for all vulcanizates, though with considerable scatter (figure 10b), which is probably related to the relatively large experimental error for TS. First TS increased with (MH-ML), then it leveled off and, finally, it decreased. Such an optimum in TS against cross-link density is normal behavior for any type of rubber, provided the crosslink density is varied over a sufficiently large range, and is due to changes in viscoelastic properties and not to changes in the intrinsic strength of the vulcanizates (ref. 25). In the same way, the tear strength showed a maximum versus (MH-ML), though with even more scatter than TS. EB showed a single correlation with (MH-ML) (figure 10c). As expected, EB decreased with increasing (MH-ML), because at higher crosslink densities, the rubber chains between the crosslinks are stretched to their maximum extent at lower strain levels. Finally, CS125[degrees]C (figure 10d) showed a very good correlation with (MH-ML). As expected, CS 125[degrees]C decreased, i.e., improved, with increasing crosslinking density.



Compound formulation [right arrow] properties

In the previous two sections, it was shown that i) (MH-ML), as a measure of the crosslink density is determined and can actually be predicted from the EPDM diene type and content and the peroxide level, and ii) the properties of the peroxidecured EPDM compounds have a single correlation with (MH-ML). Combining these two steps allows the prediction of the properties from the compound formulation. Figures 11a-c show plots of the experimental data for EB, tear and CS 125[degrees]C versus the product of the Px level and the total VNB equivalent. Good correlations were established for all three properties. Figure 11b shows that the correlation for tear was even better than the one against the experimental crosslink density (figure 4c). This signifies that establishing structure-properties relationships can significantly improve the accuracy of assessing vulcanizate properties, especially for properties that suffer from relatively large experimental errors, such as TS and tear. For HD, M300 and TS, similar plots have been obtained. These plots indicate the possibility of developing "first principle" models for predicting properties of peroxide-cured EPDM compounds.



Properties after aging

In this last section, the properties after hot air aging are considered in more detail. From the changes in properties upon aging in table 3 it was already concluded that:

* Hot air aging of these peroxide-cured compounds results in additional crosslinking; and

* aging at more elevated temperatures results in enhanced additional crosslinking.

In the previous sections, the rheometer torque difference (MH-ML) was used as a measure of the crosslink density, and correlations between the various properties and (MH-ML) were presented. Obviously, for aged samples (MH-ML) cannot be measured, thus another parameter has to be used as a measure of the crosslink density. Because M300% gives a very good fit with (MH-ML) for the peroxide-cured samples before aging (figure 10a), M300% will be used here as a measure of the cross-link density of the samples before, as well as after aging. Figure 12a shows that this approach is indeed valid, since good correlations between EB against M300% were obtained, both before and after aging (somewhat more scatter for the data aged at 150[degrees]C). Actually, the four linear fits of the data before aging and after aging at 100, 125 and 150[degrees]C fully overlap. For HD before and after aging, similar good correlations with M300% before and after aging were obtained. For TS, these correlations were inferior though, which is most probably related to the relatively large experimental error in TS, especially after aging at more extreme conditions. Figure 12b shows that aging of the 3.6% VNB-EPDM compound at more elevated temperatures resulted in enhanced crosslinking, giving an increase of M300% and a decrease of EB, simply resulting in a shift along the single EB versus M300% correlation shown in figure 12a. Similar plots were obtained for the EPDMs with different VNB contents or ENB. It is concluded that the extra crosslinks, which are formed as a result of aging, result in a network which has a similar structure and follows similar structure [left and right arrow] properties correlations to the original peroxide-cured network. Again, this is not surprising, because hot air aging results in the thermo-oxidative formation of radicals, which will induce additional crosslinking in the same way as peroxide-derived radicals. Aging simply results in additional crosslinks, affecting the properties in the same way as more crosslinks would have done as a result of higher Px levels.




The correlation between M300 and (MH-ML) for the original peroxide-cured samples can be used to calculate the crosslink density of the aged samples in terms of (MH-ML). From this it follows that (MH-ML) increased from 0.4-0.6 Nm for the original 0.2 phr Px-cured EPDM to 1.0-1.3 Nm upon hot air aging at 150[degrees]C, which is more than a doubling of the crosslink density. For the samples with higher peroxide levels, the calculated increase in (MH-ML) was less. For example, for the EPDMs cured with 2.5 phr Px, the calculated increase of (MH-ML) was only 0.1-0.3 Nm on top of 2-3 Nm, with the 3.6% VNB-EPDM being an exception.

Since M300% was shown to have a good linear correlation with the crosslink density as expressed by (MH-ML), M300% is used in yet another approach to correlate the additional crosslinking as a result of aging with the compound formulation. The difference in M300% before and after aging ([DELTA]M300%) is viewed as a measure of the additional crosslink density built up during aging. Figure 13 shows a plot of [DELTA]M300% before and after aging at 150[degrees]C against the Px level. Clearly, [DELTA]M300% decreased with increasing Px level for all EPDMs, with the exception of 3.6% VNB-EPDM. This was already observed in the results section and it was explained by the decrease in residual unsaturation upon increasing the peroxide level, due to enhanced addition reactions, and the subsequent decreased susceptibility for further crosslinking during aging. The ENB-EPDM (4.5 wt. % ENB), 1% VNB-EPDM and 2% VNB-EPDM had a similar behavior, indicating similar susceptibilities towards aging. 3.6% VNB-EPDM behaved somewhat differently, because the starting diene content was relatively high and the decrease in unsaturation upon peroxide cure was relatively small. It is concluded that a VNB-EPDM with ~2 wt. % VNB is optimal not only as far as peroxide curing efficiency and physical properties after cure are concerned, but also for aging resistance.



New Keltan ACE technology enables the production of high-molecular-weight EPDM polymers with high contents of VNB incorporated in an economically feasible manner. The peroxide curing efficiency of this new class of EPDM polymers is significantly higher, when compared with conventional ENB-EPDM polymers. In a typical peroxide-cured engine mount compound, the peroxide level can be reduced by up to 50%, while maintaining vulcanizate properties at the same level and satisfying the DBL5557 norm. This offers substantial cost savings and associated benefits, such as reduced blooming of peroxide decomposition products and reduced smell.

The aging resistance of the VNB-EPDMs is similar to that of ENB-EPDM. It is concluded that VNB-EPDM with ~2 wt. % VNB is optimal as far as peroxide curing efficiency, physical properties and aging resistance are concerned.

In a more academic approach, the EPDM polymer characteristics (in terms of the type and amount of the third monomer) and the amount of peroxide used, were correlated with the crosslink densities obtained. Next, the crosslink densities were correlated with vulcanizate properties. The results of the study indicate that quantitative models can be derived, with vulcanizate properties showing a dominant dependence on the crosslink density, which will aid the optimization of EPDM compound formulations with respect to cost and performance. Hot air aging results in additional crosslinking, but does not affect the nature of the network nor the network density [left and right arrow] properties correlations. Considering the high conversion of the VNB-EPDM unsaturation upon peroxide cure, new ways can be explored for improving the aging resistance of peroxidecured VNB-EPDMs.

(Note: Part 1 appeared in the August issue)


(1.) S. Cesca, J. Polym. Sci. Macromol., Rev. 10, 1, 1975.

(2.) G. Verstrate, Encyclopedia of Polym. Sci. and Engineering, vol. 6, 2nd Ed., 522, 1986.

(3.) F.P. Baldwin, P. Borzel, C.A. Cohen, H.S. Makowski and J.F. van de Castle, Rubber Chem. Technol., 43, 522, 1970.

(4.) H.G. Dikland, Kautsch. Gummi Kunstst., 49, 413, 1996.

(5.) H. Fujioka, Nippon Gomu Kyokashi, 51, 693, 1978.

(6.) M. van Duin and H.G. Dikland, Rubber Chem. Technol., 76, 132, 2003.

(7.) H.G. Dikland and M. van Duin, "Crosslinking of EPDM and polydiene rubbers studied by optical spectroscopy," in Spectroscopy of Rubbers and Rubbery Materials, Rapra Technology Ltd., V.M. Litvinov and P.P. De Editors., Shawbury, Shrewsbury, Shropshire, U.K., 207, 2002.

(8.) M. van Duin and B. Coussens, "(Re)evaluation of the importance of hydrogen abstraction during radical grafting of polyolefins," Polymer Processing Society meeting 11, Stuttgart 1995.

(9.) Akzo Nobel, technical bulletin on peroxide cure of rubber.

(10.) J. Fossey, D. Lefort and J. Sorba, Free Radicals in Organic Chemistry, John Wiley & Sons, Chichester 1995.

(11.) W. C. Endstra, "The application of co-agents for peroxide crosslinked EP(D)M," International Conference on Various Aspects of Ethylene-Propylene Based Polymers, Leuven, Belgium (1991).

(12.) A. Zyball, Kunststoffe, 67, 461, 1977.

(13.) D. Simunkova, R. Rado and A. Saliga, Plaste Kautschuk, 27, 247, 1980.

(14.) H.G. Dikland, L. van der Does and A. Bantjes, Rubber Chem. Technol., 66, 196, 1993.

(15.) H.G. Dikland, T. Ruardy, L. van der Does and A. Bantjes, Rubber Chem. Technol., 66, 693, 1993.

(16.) P. Windmuller and G. van Doremaele, W02005/005496 A2, 2003, to DSM.

(17.) E.G. Ijpeij, P.J.H. Windmuller, H.J. Arts, F. van der Burgt, G.H.J. van Doremaele and M.A. Zuideveld, WO 2005/090418, 2005, to DSM.

(18.) M. van Duin, M. Dees and H. Dikland, "Advantages of EPDM rubber products with VNB as third monomer, Part I: Improved peroxide curing efficiency," presented at the Fall 172nd Technical Meeting of the Rubber Division of the ACS, Cleveland, OH, October 16-18, 2007.

(19.) Rubber Technology Standards Committee, DIN 7863, 1983.

(20.) H.C. Booij, Kautsch. Gummi Kunstst., 44, 128, 1991.

(21.) Wirtschaftsverband der deutschen Kautschukindustrie, "Elastomere auf Basis Athylen-Propylen," Ed. Grunes Buch, Nr. 39, Frankfurt a/Main 1979.

(22.) W. Hofmann, Kautsch. Gummi Kunstst., 40, 308, 1987.

(23.) V.M. Litvinov and M. van Duin, Kautsch. Gummi Kunstst., 55, 460, 2002.

(24.) F.T. Wall and P.J. Flory, J. Chem. Phys., 19, 1,435, 1951.

(25.) F. Bueche and T.J. Dudek, Rubber Chem. Technol., 36, 1, 1963.

Michiel Dees and Martin van Duin, DSM Elastomers (martin.
Table 3--rheometer data and cured properties obtained
with the EPDM polymers


Px (phr) 1 2.5 5
CML (ML4) 68 67 67

Rheometer data
ts2, min. 2.0 1.2 1.0
t90, min. 7.6 6.0 5.3
MH-ML, Nm 1.3 2.4 3.5
CR, Nm/min. 0.30 0.59 0.85

Cured properties
HD, duro. A 37 40 43
M 100%, MPa 0.6 0.7 1
TS, MPa 9.8 12.9 8.3
EB, % 1,161 833 493
Tear, N/mm 19.7 12.0 4.6
CS 125[degrees]C, % 72 42 22

Aged at 125[degrees]C
TS 125[degrees]C, MPa 5.5 12.5 9.6
EB 125[degrees]C, % 1,252 842 561

Aged at 150[degrees]C
HD 150[degrees]C, A 43 44 46
TS 150[degrees]C, MPa 6.3 13.7 10.0
EB 150[degrees]C, % 1,106 854 571


Px (phr) 1 2.5 5
CML (ML4) 64 64 62

Rheometer data
ts2, min. 1.9 1.4 1.0
t90, min. 7.5 6.4 6.1
MH-ML, Nm 1.6 2.6 3.7
CR, Nm/min. 0.33 0.58 0.78

Cured properties
HD, duro. A 35 39 43
M 100%, MPa 0.6 0.8 1.1
TS, MPa 10.3 12.5 7.9
EB, % 1,078 669 399
Tear, N/mm 16.3 7.0 3.0
CS 125[degrees]C, % 63 30 16

Aged at 125[degrees]C
TS 125[degrees]C, MPa 11.6 14.0 9.0
EB 125[degrees]C, % 1,053 748 457

Aged at 150[degrees]C
HD 150[degrees]C, A 41 44 47
TS 150[degrees]C, MPa 11.8 13.4 9.0
EB 150[degrees]C, % 1,006 718 453

 3.6% VNB-EPDM

Px (phr) 1 2.5 5
CML (ML4) 60 60 60

Rheometer data
ts2, min. 1.9 1.2 1.0
t90, min. 8 6.8 6.6
MH-ML, Nm 1.6 2.9 4.2
CR, Nm/min. 0.30 0.59 0.78

Cured properties
HD, duro. A 35 41 46
M 100%, MPa 0.6 0.9 1.7
TS, MPa 11.0 9.9 6.4
EB, % 870 497 237
Tear, N/mm 9.6 3.6 2.0
CS 125[degrees]C, % 54 22 13

Aged at 125[degrees]C
TS 125[degrees]C, MPa 12.0 10.9 7.6
EB 125[degrees]C, % 839 522 285

Aged at 150[degrees]C
HD 150[degrees]C, A 40 44 48
TS 150[degrees]C, MPa 11.6 11.2 8.6
EB 150[degrees]C, % 701 459 274


Px (phr) 1 2.5 5 7.5
CML (ML4) 64 63 63 62

Rheometer data
ts2, min. 2.7 1.6 1.2 1.0
t90, min. 8.8 7.3 6.4 5.7
MH-ML, Nm 1.1 2.1 3.2 4.0
CR, Nm/min. 0.18 0.37 0.62 0.83

Cured properties
HD, duro. A 35 38 42 45
M 100%, MPa 0.6 0.7 1 1.2
TS, MPa 9.4 14.5 13.6 9.6
EB, % 1131 875 633 404
Tear, N/mm 14.6 12.4 6.4 3.6
CS 125[degrees]C, % 79 51 28 23

Aged at 125[degrees]C
TS 125[degrees]C, MPa 10.0 15.1 14.5 10.2
EB 125[degrees]C, % 1,061 862 633 401

Aged at 150[degrees]C
HD 150[degrees]C, A 39 41 43 46
TS 150[degrees]C, MPa 10.3 14.7 13.7 10.9
EB 150[degrees]C, % 1,018 841 627 447
COPYRIGHT 2008 Lippincott & Peto, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Dees, Michiel; van Duin, Martin
Publication:Rubber World
Date:Sep 1, 2008
Previous Article:Increasing fuel economy and durability in truck tires.
Next Article:Louisville hosts Rubber Division.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters