The statistical thermodynamics of defects were considered from the point of view of the thermodynamics of ideal solutions. These were then used to derive equations for the average Gibbs free energy per defect. It was noted that, when the crystal could not be treated as an ideal solution, there existed the choice of redefining the solutes so that ideality was approximately maintained or of including activity coefficients in order to account for the non-ideality. The correction terms usually changed only the square term in the mole fraction of defects. It was noted that clarification of the free energy of formation of intrinsic defects, by approaching the subject from the point of view of solution thermodynamics, permitted the general case of the simultaneous presence of any number of intrinsic defects to be treated. The higher terms in the expression for the free energy per defect could then also be easily obtained. It was pointed out that the present arguments were strictly valid only for defects (vacancies, self-interstitials, small vacancy clusters) which were not too large in comparison with the host atoms. In the case of line dislocation defects and planar defects (stacking faults, grain boundaries), it was expected that a generalization of the present argument would have to be carried out.

Average Gibbs Energy per Lattice Defect. P.T.Landsberg, S.G.Canagaratna: Physical Review B, 1997, 55[9], 5531-3