Summary

Fats and oils are, despite the hype for "light" and "diet" products, still an important ingredient to food industry. Fats can be distinguished in three different polymorphs, a, which is merely used in spreads, b', which is used for margarine and b which is important for the preparation of chocolate. Besides the polymorph used, always mixtures of different fats are used in food industry. This thesis however only contains pure fats.

After the introductory chapter, in which the reader is acquainted with the different pure fats, the present status of crystal morphology prediction models is presented. The goal of these models is to predict the crystal morphology using the internal crystallographic structure of the molecules as input. It is shown that, besides energetic criteria, also understanding of the growth mechanism of the crystals is necessary to understand the final crystal habit.

In chapters 3 and 4 are the predicted morphologies, based on the theory treated in chapter 2, compared to the experimentally observed habits. It is shown here, that for these crystals, the incorporation of the growth mechanism is essential to understand their habit. In chapter 5 the two-dimensional morphologies of two different polymorphic fats are treated. It is shown that an analogue of the morphology prediction models mentioned above can be used to understand the shapes of growth spirals and two-dimensional nuclei observed on the {001} faces of these two polymorphic fats.

Chapter 6 shows the influence of supersaturation towards the growth speed of the needle shaped fat crystals grown from different media. It is shown that the exponential to linear behavior of the growth rateversus supersaturation curves is caused a transport limitation. For solution grown crystals this limitation is due to diffusion of growth units within the solution towards the growing crystal interface itself. In the case of melt growth, impurities present in the melt have to be put forward by the growing interface, and therefore limit the growth speed.

In chapter 7 the growth shapes of fast growing paraffin crystals, which are grown using a temperature gradient are explained. Initially these shapes look like dendrites but show directly after growth large recrystallisation effects. This is exemplified by an increasing thickness of the thin platelet crystals and furthermore by the creation of holes caused by "sweating" out of solvent material. These recrystallisation effects shows a dramatic change in the shape, which makes the initially grown dendritic shape unrecognizable.

Chapter 8 and 9 describe a computer simulation model which describes the cooperation between transport and processes acting at the crystal surfaces. The Diffusion Limited Aggregation (DLA) model is used to describe the diffusion of growth units within the solution, while for the processes happening at the surface a Monte-Carlo model is used. One simulation described the crystal growth history of a confined space,
initially filled with a small seed crystal and a supersaturated solution. This model is given a thermodynamical interpretation and it shown that dissolution curves as a function of enthalpy of fusion and temperature can be simulated well. In chapter 8, the different morphologies occurring as a function of increasing supersaturation are described, which are facetted, hopper and dendritic crystals to entire networks. Here a network is defined to be a crystal which entirely fills the available space. Furthermore, it is shown that besides crystal growth, recrystallization is important, especially when the equilibrium concentration is not too small. In chapter 9 the effect of a binary system is treated, i.e. a system to which an extra growth unit, with different properties, is added. By changing the interaction between both species, mixed or separate crystals can be obtained. In some cases diffusion may cause a large difference to the composition, up to 10%, of the mixed crystals during growth.