b'Limitations of the Arrhenius Methododolgy'

The Arrhenius methodology has been utilized for many years to NOV 171998 predict polymer lifetimes in various applications. Unfortunately, there are numerous potential limitations associated with this methodology, o ST I many of which can lead to non-Arrhenius behavior. This paper will review several of these limitations, including a brief mention of diffusion-limited oxidation (DLO) effects and a more extensive discussion of the implication of changes in the effective Arrhenius activation energy E. or in the dominant reactions as the temperature changes. Changes in Ea or in the dominant reactions with temperature can happen for any material, making extrapolations beyond the experimental temperature range problematic. Unfortunately, when mechanistic changes occur, they invariably result in a reduction in effective Arrhenius activation energy, leading to lower than expected material lifetimes. Thus it is critically important to derive methods for testing the Arrhenius extrapolation assumption. One approach that we have developed involves ultrasensitive oxygen consumption measurements. Results from the application of this approach will be reviewed.