DETERMINATION OF THE EXISTENCE, VALUE AND UNCERTAINTY OF THE COMPENSATION OR ISOKINETIC TEMPERATURE

The demonstration of the existence of a compensation effect, the linear dependence of the enthalpy on the entropy for a series of chemical reactions, is often complicated by an apparent linear dependence arising solely from experimental errors. A non-linear least squares method has been developed for establishing the existence or non-existence of a compensation effect and for determining the value of the compensation temperature if the effect exists. The uncertainty in the value of the compensation temperature can be determined by means of a Monte Carlo procedure. If a compensation effect does no exist, three cases can be distinguished: (1) the relationship between the enthalpy and entropy is linear, but the compensation temperature is 0 K; (2) the various reactions cannot be distinguished from each other because of experimental errors; and (3) the reactions do differ, but the relationship (if one exists) between the enthalpy and entropy is not linear. The method is illustrated using both hypothetical and experimental examples.