Bias in isothermal time-to-event studies due to approach to test temperature

Abstract

Oxidative induction time (OIT), constant temperature stability (CTS) and isothermal crystallization are examples of isothermal time-to-event (TTE) measurements obtained using differential scanning calorimetry. In TTE experiments, a test specimen is heated/cooled at a constant rate from the setup temperature to an isothermal test temperature. Once the test temperature is achieved, a clock is started and the time to the thermal event (e.g., onset to oxidation, thermal decomposition or crystallization exotherm peak) is measured. Such TTE values may be used to rank stability of the material at the test temperature. Some portion of the reaction of interest, however, takes place during the pre-isothermal period as the test specimen approaches the test temperature. This amount of reaction is unmeasured and represents a bias in the resultant TTE value. An equation has been derived and numerically integrated to estimate this bias. This approach shows that the bias is dependent upon the activation energy of the test reaction, the heating/cooling rate used and the temperature range between the melting temperature and the test temperature. For commonly used heating rates, the bias for OIT and CTS tests is small. Further, the myth that isothermal crystallization kinetics determinations required high cooling rates is dispelled with the bias of less than 0.9 min resulting from heating rates as low as 10°C min⁻¹. Knowledge of magnitude of this bias permits the selection of experimental conditions without the expense of high heating/cooling rate apparatus or extra cost cooling accessories.