Optically and chemically zoned white micas are a major mineral phase in a small high-level microgranite body (Argemela, Central Portugal). Contacts between dioctahedral core (phengite) and trioctahedral rim (trilithionite) are sharp but irregular in detail. This suggests disequilibrium and topotactic replacement of the primary phengitic mica which became unstable in the presence of a Li-(Rb, Cs)-F-rich environment. In-situ analyses show a very good linear correlation between Li and F in the trilithionite rim. Such a strong Li and F enrichment from core to rim is very comparable with that commonly observed in chemically contrasted mica generations from outer and inner parts of evolved pegmatites, respectively, where an inward progressive crystallization sequence is often postulated. Raman scattering and infrared absorption spectra from core and rim are nearly identical and characteristic of the muscovite (dioctahedral) species. It is likely that Li and F belong to a trioctahedral environment, whereas OH remaining after F substitution belongs to a relictual dioctahedral environment. This explains why the strong OH-stretching vibrations in the high-frequency range (around 3650 cm(-1)) are similar whatever the nature of the mica, muscovite or trilithionite. The remaining OH vector in the lepidolite rim lies close to the cleavage plane, as in muscovite. According to their contrasted chemical compositions, structural Variations between core and rim might be expected. However, X-ray powder diffraction and (MAS) NMR spectra obtained on a composite mica separate have strong similarities with those obtained for comparison on a pure phengite separate, probably by lack of sufficient resolution. This shows that results obtained from average compositions can be partly deceptive.
