The speckled nature of all the crystals is in sharp contrast to that seen during tripalmitin crystallisation, where well formed circular crystals were formed with no internal cavities.
This is a consequence of crystallising from a pure melt. With palm oil at these temperatures, a significant “low melting” fraction exists which is unable to crystallise and must diffuse away from the interface.
It is apparent from the morphologies produced that much of this uncrystallisable “solvent” fraction is unable to diffuse away fast enough to avoid becoming trapped by crystal growth around it – giving rise to the speckly morphology.
In the microscope slide environment there are likely to be few convection currents and so mass transfer would be almost purely by diffusion.
The situation, however, is likely to be different in a stirred tank or other flow environment where the “solvent” molecules may have a better chance to be transported away from the crystal surface before becoming entrapped.
The non-circular nature of some crystals (see Fig. 2d) also suggests that non-uniformities in composition of the liquid phase arise due to differences in the extent of crystallisation locally, giving different growth rates in different directions. So the liquid phase cannot be considered to be fully mixed in these experiments.