The variations of the dimensionless concentration gradients averaged over each of the three different faces of the growing crystal (left <001>, right <001> and base <100>) with respect to experimental run time are shown in Figure 5.22(a) for 0 and 15 rpm. As expected, left-right symmetry of the crystal is realized in growth with and without rotation. The effect of rotation is to lower the overall concentration gradient when compared to that generated by buoyancy alone. The gradients on the lower face are small in comparison to the sides. On the lower face, density stratification is stable and buoyant motion is inhibited. The effect of rotation is then to increase the gradients here by inducing fluid movement. Evolutionary profiles on the top face of the crystal are not shown, since the image is adversely influenced by the presence of the seed holder.

 In the purely buoyancy-driven experiment, the gradients on the side faces grow in strength with time, whereas the gradients along the lower face are small. The increase in the gradients on the side faces is consistent with the corresponding high intensity regions in the schlieren image sequence shown in Figure 5.19. The problem of high concentration gradients during the later stages of experiments (t = 60 hours) and also a significant difference in the relative distribution of these gradients over sides and lower faces of the growing crystal is seen to be overcome by rotation. The effect of rotation in equalizing the strength of the gradients over the three faces of the crystal is indicated by the proximity of the gradient profiles in Figure 5.22(a). Figure 5.22(b) shows the horizontal growth of the crystal with respect to the experimental run time. The growth rate with rotation is slightly lower when compared to that based on buoyancy alone; it is however practically linear. The growth rate with crystal rotation is comparatively lower because of two factors: (a) the lowering of concentration gradients in the vicinity of the growing crystal due to homogenization of the solution induced by crystal rotation, and (b) the rotation of the crystal introduces a radial (outward) velocity component that inhibits the transport of solute to its growing surfaces. Figure 5.22(c) shows photographs of the grown crystals for the two cases. The size of the finally grown crystal (after 90 hours) is larger in buoyancy-driven convection, but the crystal quality is superior in terms of transparency when growth is accompanied by rotation.