Title thesis: Crystal growth studied on a micrometer scale

Ir. M. Plomp

Promotor: Prof. Dr. E. Vlieg en Prof. Dr. P. Bennema

Co-promotor: Dr. W.J.P. van Enckevort

KU NIJMEGEN 13/12/1999

This thesis is a collection of eleven crystal growth studies, which have their subject and the observation technique used in common. All of them deal with crystal faces, which are studied mostly with the help of atomic force microscopy (AFM).

The basic question that the science of crystal growth tries to answer is: How does a large amount of separate building blocks (atoms, ions, molecules) evolve into a macroscopic crystal with a number of specific crystal faces? The research presented in this thesis is addressed to one aspect of this general question: How does one particular crystal face grow?

In many cases, crystal faces are not roughened en growth proceeds layer by layer, in which every new layer of growth units spreads over the surface via steps. In situ and ex situ AFM provides a relatively easy way to image these monomolecular step patterns. The high resolution of AFM even enables direct imaging of the molecular surface lattice.

Within the general resemblances mentioned above, there is a large diversity of the subjects in this thesis. Most eye-striking is the difference in investigated crystals: ionic crystals on one hand, and organic crystals (paraffins, fats and proteins) on the other hand. In addition, the topic of the various crystal surface studies also varies: crystal growth, crystal dissolution, cleavage, the influence of the AFM itself on the crystals.

The first two chapters treat the growth mechanisms of barium nitrate crystal faces. The presence of several types of screw dislocations, leading to different types of growth spirals on the {111} and {100} crystal faces, as well as the occurrence of 2D nucleation on these faces, is elaborated in chapter 1. Chapter 2 deals with the occurrence and the size of the hollow cores that can emerge at the spiral centres. The size depends on both the dislocation's burgers vector magnitude and the supersaturation during growth. This behaviour is predicted by theoretical calculations, which seem to hold remarkably well for the small burgers vectors involved.

The third chapter addresses the growth of silver bromide tabular crystals, which are important for the photographic industry. Several ex situ and in situ AFM observations indicate that these crystals grow with layers of one AgBr growth unit. Furthermore, observations of nucleated crystals confirm the occurrence of twinning, which is the key mechanism for the growth of tabular crystals.

The next four chapters are all about potassium bichromate (KBC). This crystal has many interesting properties, such as the opposite (001) and (00-1) crystal faces that grow completely different, despite the crystal symmetry as determined by X-ray diffraction. Further, the crystal structure is composed of alternating ionic layers A and B parallel to {001}. In the first KBC chapter the growth of the deviating (00-1) face is studied. It appears that the steps of one or a few unit cells in height are relatively easily blocked, while macrosteps of hundreds or even tens of thousands unit cells in height can still grow, and will develop to micro-facets. This phenomenon often leads to the formation of 'caves' in the crystals, which is attributed to the combination of impurity blocking and volume diffusion effects. In the second KBC chapter the double layered structure perpendicular to <001> is treated in more detail. In case of cleavage of a crystal along {001}, the cleavage plane is between layers A and B or between B and A, depending on the cleavage direction, as a result of the crystal symmetry. From these experiments, together with the KBC dissolution experiments described in chapter 7, it is proven that the A layer is thermodynamically the most stable layer, and is almost always on top of both opposite {001} crystal faces. When freshly cleaved crystals are exposed to air, the upper B layer that is present on one crystal half dissolves and recrystallizes as A. This process is facilitated by the ultrathin water layer that forms on most crystals exposed to air. This water layer is also detected by the use of tapping mode AFM (TM-AFM) on KBC crystals exposed to air, as is demonstrated in chapter 8. The water layer introduces a large attractive capillary force between the TM-AFM tip and the crystal, and this again induces mode switching between an attractive and a partly attractive, partly repulsive mode. This results in an AFM image having holes of a few nm depth, which correspond to these mode switches rather than to real topographic height differences.

The first part of the thesis deals with ionic crystals, while in the second part some crystals of organic substances are studied. These molecules form much larger growth units than ions. The Van der Waals bonds keeping the crystals together are relatively small, which makes the crystals softer than in the ionic case.

The first 'organic' chapter deals with the spontaneous movement of screw dislocations in paraffin crystals after cessation of growth. If this happens, a screw dislocation leaves a capricious step as a trace, that can intersect the straight spiral steps formed during growth and even itself. The movements are induced by shear strains, which probably are caused by adhesion between thin crystal sheets. In chapter 10 paraffin crystals with a melting point that is relatively low, but still well above room temperature, are studied by means of TM-AFM. It appears that the AFM tip, which is heated by a focused laser beam in order to measure its deflection, locally melts the crystal surface. A liquid bridge between tip and sample evolves, and transport of liquid paraffin molecules towards or from the tip takes place, which corresponds to etching or growth of the crystal surface, respectively. In this way, melt growth of paraffin can be studied 'in situ' by AFM.

In chapter 11, the relation between crystal structure and surface morphology is examined in a detailed way. As an example, the morphology of growth spirals on fat crystals is compared with the surface morphology as deduced from the crystal structure by different methods. It follows that the combination of Hartman Perdok theory and step kinetics is the best method to predict two-dimensional morphology.

The final chapter consists of an in situ AFM investigation of the growth of insulin protein crystals. Three types of growth were discerned: 2D nucleation, spiral growth and a form of 3D nucleation. If screw dislocations are present in the crystal, spiral growth dominates at the low supersaturations applied. If these are absent, the crystals are forced to grow by 2D nucleation. Occasionally, large aggregates or crystallites land on the mother crystal and grow out laterally. Only a weak overlap of the diffusion fields of the steps was recorded.