Matskási István (szerk.): A Magyar Természettudományi Múzeum évkönyve 82. (Budapest 1990)

Papp, G.: A review of the multi-layer lizardite polytypes

MELLINI (1982) and MELLINI & ZANAZZI (1987) found lizardite specimens to be suitable for three-dimensional structural analysis. They refined the crystal structure of lizardite-lX to R=0.031 and 0.074 (other sample) and of lizardite-2H

1 to R=0.024. The crystal structure of the multi-layer lizardite polytypes is not known in details. Nevertheless, the polytypism of the trioctahedral 1:1 layer silicates was discussed by several authors (references are given by WICKS & WHITTAKER 1975, or BRINDLEY & BROWN 1982). Lizardite polytypes and their classification" Theoretical models assume ideal structure without distortion. Successive 1:1 layers are stacked in such a way that hydrogen bonds connect them. The layer stacking is optimal for the hydrogen bonds if successive layers are shifted relative to one another by (i) a/3 along the three possible X. axes (or -a/3 if Mg occupies the position labelled I by BAILEY 1969), (ii) ±b_/3 along the three possible Y_ axes; or (iii) the layers are rotated around 1 by +60°, +120 , or 180 . The combination of these operations produces a lot of possible stacking sequences even if the symmetrically equivalent combinations are not considered. BAILEY (1969) derived 12 standard polytypes (Table 1) using the limiting assumption that only the same kind of shift (a/3, ±b_/3 or no shift) combined with the same kind of rotation (±60 or no rotation) may occur within a given polytype

. The identification of the actual polytype is usually made through the evaluation of X-ray precession photographs of single crystals or X-ray powder patterns. Single crystals of lizardite occur very rarely, therefore we are confined to the latter. On the basis of the j<=3n_ type reflections (orthorhombic or monoclinic setting) the mineral can be classified into one of the four groups A-D, whereas the identification of the individual polytype is based on the k_*3n_ type reflections (for details see BAILEY 1969, 1988). The weakness of k/3n_ type reflections and the disordered layer stacking usually prevent us from identifying of the actual polytype. The stacking sequence was determined only in the case of Unst serpentine so far. (It is of some interest that the polytype sequence of this mineral is not included to the 12 "standard" polytypes . ) A new classification of multi-layer Mg-serpentines Classification of multi-layer lizardite polytypes is based on X-ray investigations as discussed above. Transmission electron microscopy (TEM) was used only for obtaining some additional data mainly about particle morphology. The development of TEM methods and the discovery of polygonal serpentine allow us to consider this mineral group from a different point of view. Roentgenographically multi-layer lizardites can be divided into two groups on the basis of TEM investigations. The first group comprises the minerals proved to be lathy ("L" in Table 2) when viewing their crushed samples by TEM. These specimens are in all probability not lizardites sensu stricto

, but they are composed of polygonal serpentine in greatest proportion. In the second group can be classed the minerals forming (pseudo) hexagonal tabular crystals ("T" in Table 2) of microscopic or submicroscopic size. Only these specimens are multi-layer lizardites in a proper sense. From pure structural point of view (that is regarding only the layer stacking) there is no sharp difference between the two groups. For example, HALL et al. (1976) determined the polytype sequence of Unst serpentine ("L" group) using the X-ray data of better formed Mg-Ge-"serpentine" with the same layer stacking ("T" group). Multi - layer polygonal' serpentine - Polygonal serpentine was recognized not long ago (a summary is given by PAPP 1988, or WICKS & O'HANLEY 1988). This peculiar variety of serpentine displays lizardite-, chrysotile-, and antigorite-like features almost equally. Polygonal serpentine is composed of flat layers arranged polygonally around the common X axis. The polgygonal "rod" is built from 30 (or 15) sectors of