Evidence of non-Arrhenius behaviour from laboratory aging and 24-year field aging of polychloroprene rubber materials

https://doi.org/10.1016/j.polymdegradstab.2004.06.010Get rights and content

Abstract

High-temperature oven aging exposures from 110 °C to 80 °C have been conducted on a commercial chloroprene rubber cable jacketing material where the time-dependent degradation at each temperature is monitored by following the ultimate tensile elongation. By time–temperature superposing the results at the lowest experimental temperature (80 °C), empirical shift factors are first derived and then analysed using the conventional Arrhenius approach and finally extrapolated using the derived activation energy Ea of 96 kJ/mol in order to make predictions at 25 °C. To test the constant Ea assumption underlying the Arrhenius extrapolation, we conducted oxygen consumption measurements at six temperatures ranging from 25 °C to 95 °C and found evidence for a slow drop in Ea from ∼96 kJ/mol above 80 °C to an average value of ∼82 kJ/mol below 80 °C. This curvature predicts a 50% reduction in room temperature lifetime and suggests that certain higher Ea degradation processes become less important as room temperature is approached. Analyses of the production rates for CO2 show that CO2 becomes less important relative to the oxygen consumed as the temperature is lowered, evidence in accord with the above suggestion. Further evidence for a similar drop in Ea (from 89 kJ/mol above 70 °C to 71 kJ/mol below 70 °C) comes from comparisons of accelerated aging elongation data taken ∼25 years ago on another chloroprene rubber cable jacketing material with recent elongation results taken on samples that have aged for ∼24 years at ∼24 °C.

Introduction

An extremely important objective and a significant challenge for the polymer community has been the prediction of polymer service lifetimes. This is especially important for certain materials such as those used in nuclear power plant safety cables since these materials may be called upon to operate reliably in high level accident environments after ambient aging for upwards of 40 years [1]. Historically, the so-called Arrhenius approach has been utilized to analyse shorter-term high-temperature accelerated aging exposures and then to extrapolate these results to make long-term predictions at the much lower ambient temperature of interest. This approach is based on the assumption that the degradation involves an activated chemical process whose rate is proportional to exp(−Ea/RT) where Ea is the Arrhenius activation energy, R is the gas constant and T is the absolute temperature. Once an Ea value is obtained empirically from the accelerated experimental results at high temperatures, it is used to make extrapolated predictions at lower temperatures with the assumption that Ea remains constant over the extrapolated temperature range.

We have been interested for many years in developing better methods for analysing and extrapolating accelerated aging results [2], [3], [4], [5], [6], [7]. Since oxidation reactions typically dominate degradation in oxygen-containing environments (e.g., air), one approach that has been found to be particularly useful for testing the Arrhenius extrapolation assumption is based on ultra-sensitive oxygen consumption measurements [3], [4]. We take these measurements at high temperatures overlapping traditional accelerated aging conditions to confirm that the activation energies derived from the high temperature oxygen consumption are consistent with the activation energy derived from traditional mechanical property measurements. In addition, the sensitivity of the oxygen consumption approach allows us to probe much lower temperatures allowing direct measurements over the normally inaccessible extrapolation regime. This means we can determine whether the activation energy for oxidation changes in this region thereby quantitatively testing the fundamental Arrhenius assumption.

We have now studied numerous materials utilizing this approach including a nitrile rubber [3], [8], a neoprene seal material [3], [9], an EPDM seal material [4], a polyurethane binder material [10] and two butyl seal materials [7]. In every instance, the high temperature Ea for oxygen consumption is in excellent agreement with the Ea found for conventional measurements (mechanical properties, density, growth in infrared signals) at high temperatures. In the lower-temperature extrapolation region, however, curvature to lower values of Ea is often observed [4], [7], [9], [10] implying that the assumption underlying the Arrhenius approach may overestimate material lifetimes. Such curvature to lower Ea values as the temperature is lowered (often referred to as “downward curvature”) is not unexpected since a degradation pathway involving a combination of a high Ea process in parallel with a low Ea process will favor the low Ea process as the temperature is reduced. Evidence for this downward curvature has been observed numerous times in the past. An early example is work by Richters on polypropylene where he found curvature to lower Ea values for the oxidative induction period [11]. Gugumus reviewed the pitfalls of extrapolation [12] and Bernstein and Lee showed evidence for downward curvature in Arrhenius plots for oven aging of HDPE insulated cables [13]. Gijsman and co-workers found downward curvature for unstabilised polypropylene and suggested an explanation based on changes in the hydroperoxide decomposition mechanism [14]. Rosik and co-workers showed similar failure of Arrhenius extrapolations of oxidation induction time results for a stabilized ABS polymer concluding that the cause was a change in the mechanism of oxidation reactions [15]. Gugumus recently presented evidence of downward curvature for unstabilised polypropylene films but found a second upward curvature at aging temperatures below 80 °C for films stabilized with phenolic antioxidants [16]. Although this upward curvature could not be explained, it was tentatively attributed to some kind of complexation of titanium by the phenols leading to a possible passivation of the transition metal. This observation of upward curvature is quite unusual, ignoring of course anomalous upward curvature effects caused by diffusion-limited oxidation [6].

In this paper, we apply the ultra-sensitive oxygen consumption approach to a commercial chloroprene rubber cable jacket and find evidence of a drop to lower values of Ea in the extrapolation region. We also present results showing that the production of CO2 becomes less important as the temperature is reduced, consistent with the picture that a changing mix of degradation reactions must underlie curvature in the Arrhenius curve. In addition, we present elongation results for another chloroprene rubber cable jacketing material that has been exposed to ∼24 years of ambient (∼24 °C) aging. Time–temperature superposition of the elongation results from previous accelerated aging experiments at temperatures ranging from 70 °C to 131 °C with the 24-year results at 24 °C leads to downward curvature in the Arrhenius plot similar to that found for the other chloroprene rubber material.

Section snippets

Materials

The two chloroprene rubber cable jacket materials are from commercially produced low-voltage cables that are qualified for nuclear power plant safety applications. The first (referred to subsequently as CR-1) was manufactured by the Rockbestos Cable Co. in 1989 and represents the jacket removed from a 600V Firewall III cable which contained XLPE insulations surrounded by a chloroprene rubber jacket of average thickness ∼0.15 cm. The second (CR-2 from now on) was the cable jacket (0.15 cm nominal

Conventional extrapolation of tensile elongation results for CR-1

Fig. 1 shows the tensile elongation results for CR-1 as a function of aging time at 80 °C, 95 °C and 110 °C plotted versus log of the aging time. Since complete loss of tensile elongation takes on the order of 2 years at 80 °C, little useful information on mechanical degradation would be typically available for aging carried out below this temperature. A typical Arrhenius analysis of these data would select a “failure” criterion such as 50% loss of initial elongation and plot the times to this

Conclusions

For two chloroprene rubber cable jacketing materials, we have shown evidence for non-Arrhenius degradation behaviour. For the first material (CR-1) elongation and oxygen consumption results indicated an Arrhenius activation energy of 96 kJ/mol above 80 °C; lower temperature oxygen consumption results showed that the activation energy decreased to an average value of 82 kJ/mol below 80 °C. For the second material (CR-2), elongation measurements taken on samples that had aged at 24 °C for 19.2 and

Acknowledgements

Sandia is a multiprogram laboratory operated by Sandia Corporation, a Lockheed Martin Company, for the United States Department of Energy's National Nuclear Security Administration under Contract DE-AC04-94AL85000. The current studies were part of the Nuclear Energy Plant Optimization (NEPO) program jointly funded by the DOE and by the Electric Power Research Institute (EPRI). The authors appreciate the competent technical assistance provided by Mike Malone and Michelle Shedd on the elongation

References (27)

  • Gazdzinski RF, Denny WM, Toman GJ, Butwin RT. Aging management guideline for commercial nuclear power plants—electrical...
  • K.T. Gillen et al.

    Accelerated aging methods for predicting long-term mechanical performance of polymers

  • K.T. Gillen et al.

    Extrapolation of accelerated aging data—Arrhenius or erroneous?

    Trends Polym Sci

    (1997)
  • Cited by (133)

    View all citing articles on Scopus
    View full text