Interplay of impurities and solution flow as determinants of step pattern dynamics.

The first theory of step pattern evolution, the kinematic wave theory, employed the assumption of impurity effects on step kinetics. On the other hand, recent results have been considered within a framework linking step patterns to the mutual orientation of the solution flow and step motion directions in arbitrarily pure solutions. We explore the consequences of combining impurity and solution flow effects on the dynamics of the surface morphology of the (101) face of potassium dihydrogen phosphate (KDP) crystals. We employ phase-shifting interferometry for real time in situ monitoring of these dynamics. We find that, at solution supersaturations sigma0.040 impurities do not affect step bunching, and it is controlled by the direction of the solution flow, i.e., two distinct regimes of step bunching exist. The transition between the two regimes is governed by the exposure times of the terraces between steps tau : shorter tau's at higher growth rates lead to lower surface concentration of impurities and suppress the impurity effects on step kinetics and bunching.

Step bunching in a diffusion-controlled system: phase-shifting interferometry investigation of ferritin.

We present a novel phase-shifting interferometry technique for investigations of the unsteady kinetics and the formation of spatio-temporal patterns during the protein crystallization. We applied this technique to the ferritin crystal growth, which is controlled by the rate of supply of material. We find strong fluctuations of growth rate, step density and step velocity due to passage of step bunches. The fluctuation amplitudes decrease with higher supersaturation and larger crystal size, as well as with increasing distance from the step sources. Since these are parameters affecting the solute supply field, we conclude that fluctuations are rooted in the coupling of the interfacial processes of growth to the bulk transport in the solution. Analysis of the step velocity dependence on local slope indicates a very weak interaction between the steps. Hence, in diffusion-controlled systems with non-interacting or weakly interacting steps the stable growth mode is that via equidistant step trains, and randomly arising step bunches decay.