Melting curves (temperature‐heat content curves) as criteria for purity

Abstract

AbstractMethods for calculating the purity of a given sample from its melting curve developed thus far have been derived by assuming that the contaminants are immiscible with the main component in the solid phase. These methods may give rise to serious errors when applied to arbitrary samples.In the present paper a discussion is given of the behaviour of systems with solid solutions (miscibility in the solid phase), and a method is developed, resembling Wichers' “comparative method”, that is applicable to both types of systems.In addition, errors occurring in the experimental determination of melting curves are briefly discussed.Cooling curves are considered to be less suitable for the determination of purity than heating curves.A description of a simple apparatus and procedure for the determination of heating curves is given.The quantity of substance needed for a determination is about 200 mg. It is shown that the smallest amount of contaminant (s) determinable with this apparatus varies (depending on the system investigated) from 0.1 to 0.01 mol. %. The sensitivity of the method may be improved by using more sensitive thermometers.