Growth from Liquid or Vapour

There are some comments to make here. First, the positions of the Sn, Pb, or Ba atoms in these phases is not known, so that for these compounds the model must be regarded as hypothetical. Secondly, it does not suit needle-shaped crystals, as the first layer laid down must already include the fault planes inherently contained within it. This point will be returned to later. Thus the alkali-metal intergrowth phases, which are needle shaped, do not fall into this class. This growth mechanism can also be applied to the barium ferrites, which are plate-like crystals that form from a molten flux, and it is possible that it could account for the sort of faulting noted in the amphibole chain minerals described on p. 135 et seq. 

Here the major question to be answered is whether the various polytypes of a particular phase do indeed have identical compositions layer for layer. K. H. Jack 

Other illustrations could be given. However, we have taken this model far enough for the present purposes, and ultimately it should be regarded as hypothetical. Careful comparison with experiment is needed to elucidate matters further. The purpose of the preceding discussion is simply to point out that interactions between fault planes and elaborate solid-state rearrangements involving a flux of point defects are not necessary in order to produce some of the seemingly complex structural sequences illustrated in this review. 

When needle-shaped phases, such as the W 03 2 CS structures are considered, the fault planes run parallel to the needle axis. In addition, the needles are rarely homogeneous, and often consist of domains in which the planar defects are relatively well ordered, but oriented differently from domain to domain. It is this feature which has made these crystals so difficult to study by A -ray diffraction. No mechanism has been proposed in the literature to account for the formation of such materials, and so they will not be considered further here. 

---