Ti- and Ba-rich tri-octahedral micas occur in fractionated basic igneous rocks, metasomatized mantle peridotites, metamorphosed pelites/carbonates, and hydrothermally altered mineral deposits. Electron microprobe analyses (EMP), with all iron reported as FeO, were widely used in the 1970/80s to interpret Ti and Ba substitution mechanisms, based on 22 O2– unit cell calculations, implying that cation vacancies occur in octahedral and/or intersheet sites. In 1996 EMP with chemical and physical analyses for ferric and total Fe, H2O, (OH), and element-specific Fe X-ray Absorption Spectroscopy (both K and L-edges) established valence states for Fe and Ti and cation site occupancies, that ∼50 % O replaces (OH) molecules, and that 24 anion cell formulae show the absence of cation vacancies. Cell formula calculation protocol for phlogopitic micas is refined here and results tested against the stoichiometric formula for vacancy-free phlogopite, XIIK2VIMg6IV[Si6Al2]O20(OH)4. Hypothetical sheet silicate compositions, calculated with fixed contents of vacancies linked to particular mixed-valence element substitutions, confirm that reliable unit cell formulae for natural mica solids require that each stoichiometric vacancy must be accounted for. If reliable estimates for ‘excess O’ (denoted WO2−) are assigned to EMP analyses, the proportion of the oxy-mica component in a mica solid solution can be defined. This approach is tested using published analyses for Ti- and Ba-rich biotites from fractionated basic and ultramafic volcanic igneous rocks (oxymica range 2.5–45 %; TiO2 up to 14 %; BaO up to 23 %), upper mantle peridotites (equivalent values 7–18 %; 6 %; 0.7 %), and metasomatised upper mantle (2–37 %; 9 %; 23 %). Enrichments of Ti and Ba in micas are clearly linked to the extra oxygen charge required to neutralise the more highly charged Ba2+ and Ti4+ replacing K+ and Mg2+.Substitution mechanisms involving Ba, Ti, and Fe3+ in ideal phlogopite involve coupled inter-site/inter-valence interactions so both site-ordering and valence-balance between the different cation sites must be accounted for. The cation exchange 2XIIK+ + 4VI(Mg2+ + Fe2+) + 4IVSi4+ ↔ XIIBa2+ + 3VITi4+ + 4(Altotal + Fe3+ + Crtotal)3+ is used here so that valence and site order can be considered together. Compositional variations show well-defined trends towards high-oxymica content and vacancy-free Ti and Ba mica end-members, rather than to oxy-free end-members with essential vacancies. Average compositions for different source rock micas are used to assess compositional trends; molecular WO2− values are refined by relating initial estimates directly to the wt.% TiO2 contents; and refined WO2− and stoichiometrically calculated Fe3+/Fetotal ratios ranging ∼0.1–0.5) are correlated with possible primary magmatic values.
Read full abstract