A linkage between the activity and the chemical potential

A minimisation of an appropriate characteristic function (such as the Gibbs energy) of a system resulting in establishing its equilibrium state does not require a notion of component's activity, which, in contrast to the chemical potential, is nothing more than a helpful auxiliary entity bounded by finite limits. Also, the minimisation procedure does not impose rules on how a reference state for each component should be chosen, because the equilibrium state is not affected by particularities of such a selection. However, the choice of the reference state and the definition of activity become important when it is needed to determine what will happen with a numerical value of the activity of a component in a phase (such as molten steel or slag) when a reference state with respect to which it was specified changes. In this paper, it is argued that the activity ought to be introduced as rather than .