Crystal growth science between the centuries

The major framework of crystal growth concepts and technologies was built during the 20th century. This solid framework, however, does not have enough predictive power. The lack of fundamental constants, multi-parametric experimental conditions, and complex chemistry in many systems remain obstacles to the quantitative confrontation of many key concepts with experiment, to make them predictive and to develop these concepts further. In addition, at least several generic issues should be addressed. For instance, the BCF theory is based on the assumption that thermodynamic fluctuations at steps are fast enough to generate a sufficient density of kinks for growth. However, for strongly polygonized steps this is not always the case. Also, for these steps, the Gibbs-Thomson relationship may be valid only within a very low supersaturation range. These problems are considered here. The Kossel model should be generalized for the lattices of which unit cells include several, say, n identical molecules, atoms, or ions in crystallographically inequivalent positions. In this case, only the unit cell as a whole makes a self-reproducible kink while the building blocks in the liquid or gas are the component molecules, atoms or ions. Because of this, the driving force for crystallization takes the same form as that in a system of n components. Therefore even the kink rate is proportional to C-n-C-e(n) rather than to C-C-e where C and C-e are actual and equilibrium concentrations, respectively. Finally, the challenges of biomacromolecular crystallization are discussed. (C) 2001 Kluwer Academic Publishers.