Liquid-phase precipitation nucleation kinetics

An interesting difference between the samples is found in the behaviour of the samples during destabilisation. While for the sample under a deuterium pressure, LiBD4 is not destabilised until both phases are molten, in the vacuum sample LiD is destabilised in the solid phase, and at much lower temperatures (ca. 360°C). This is primarily due to the improved diffusion kinetics of Li species over that of LiBD4. There is also the possibility that LiD precipitation out of the liquid phase at nucleation sites on the Mg particles allows improved mixing over the solid Mg and liquid LiBD4 observed in the sealed sample. 

When we consider the formation of the MS (which we assume to be the closest to the sol phase), then the free energy gap AGy is lower than the most stable phase, referred to as SS. Thus, according to CNT (Equation 1.2), the only way that the phase MS can precipitate out at any temperature is to have such a lower surface tension that the nucleation barrier is lowest for MS. Because of lower solid-liquid surface tension leading to a lower nucleation barrier, MSs are kinetically favored. Thus, kinetics seems to play a very dominant role. 

---