When is a compound energy not a compound energy? A critique of the 2-sublattice order/disorder model

In recent years an extension of the 2-sublattice model has been used to model ordering transformations in systems such as Ni-Al and Ti-Al. The model provides a means by which the Gibbs energy of both the ordered and disordered phases can be represented using a 'single' equation thus allowing a continuous change in Gibbs energy to be obtained between the ordered and disordered states. This is an attractive proposition as it permits second order transformations to be considered within the scope of readily accessible models in CALPHAD software packages. However, it can be demonstrated that there are some fundamental flaws in the model, particularly when asymmetric ordering energies are used. This leads to the existence of a residual, extraneous Gibbs energy in the disordered state originating from the ordering parameters. In the most recent expositions of the 2-sublattice order/disorder model the residual energy is empirically removed but this leads to the position where it is possible that the Gibbs energy which is actually calculated for the fully ordered state;of a compound such as an AB or A(3)B type to be quite different from that actually specified for G(AB)(ord) or G(A3B)(ord) hence the title of the present paper.