Abstract: Hydrothermally kaolinitized muscovite from the Otago schist of Brighton, New Zealand, has been studied by transmission and analytical electron microscopy (TEM and AEM) to determine the mechanism of alteration and to compare reactant-product relations for di- and trioctahedral micas. The muscovite is a primary metamorphic phase having a phengitic composition. It occurs as well-ordered, two- and three-layer polytypes, in grains as thick as 30 µm. Kaolinite occurs as packets of layers, each about 100–600 Å thick, which alternate with packets of muscovite or smectite-like layers. Most of the kaolinite is highly disordered in stacking sequence, although a one-layer polytype is also present, occurring as relatively thick sequences of layers. Phase boundaries between kaolinite and muscovite are invariably parallel to the 001 lattice fringes with no strain contrast; i.e., no transitions exist along layers. Parallelism of 00l and 11l reflection rows of both kaolinite and muscovite implies a topotaxial intergrowth. A smectite-like phase is also present, occurring as packets of wavy layers, which locally have periodic contrast that may reflect R1 ordering of illite/smectite. This material appears to be a direct, “along-layer” alteration product of muscovite. Electron diffraction data and lattice-fringe images imply that kaolinite alternates with micaceous phase(s) with some regularity; i.e., micaceous layers are separated by approximately equal numbers of kaolinite layers. Similar long-range periodicity occurs in contrast variations within packets of kaolinite layers.
The data collectively suggest that the alteration interface was self-perpetuating and that alteration proceeded rapidly along layers once it initiated in 2:1 layers at crystal edges or strained areas, with no observable component normal to the layers. They also suggest that smectite may have formed as an intermediate phase during the hydrothermal kaolinitization of muscovite. In the previous study of alteration of biotite in the same sample, “along-layer” transition boundaries were commonly observed, and a second, intermediate product phase was not detected, implying a relation between the alteration mechanisms and the chemical differences between reactants and products.