Diagenetic clay minerals usually occur as heterogeneous assemblages of submicroscopic layers consisting of different structure types such as illite, smectite and chlorite, with variable composition within a given structure type, and with highly variable concentrations of imperfections. The dimensions of mixed-layering, the semi-coherent to coherent nature of the structures across the layering, and compositional heterogeneity occur at a scale well below that of an individual thermodynamic phase. These relations imply that most clays are not distinct minerals or phases, and that assemblages of clays in shales and mudstones are incompatible with the phase rule. Such relations are better evaluated in terms of the formation of metastable materials with each small unit having unique chemical properties, rather than as a small number of stable homogeneous phases. Consequently, treatment of most clay minerals in terms of equilibrium stability with either a thermodynamic or experimental approach is subject to error.

Chemical reactions involving most clay minerals are best understood with kinetic models. These involve a great variety of parameters such as time, fluid/rock ratio, deformation history, nature of starting materials and transformation mechanisms, as well as the variables, such as temperature, pressure and composition, that are commonly used to define equilibrium. Solubility experiments on the stabilities of clay minerals are unlikely to attain equilibrium at low temperatures. Moreover, the activity of soluble species may be controlled by surface equilibria, or by absorbed or exchangeable cations. Interpretations of available experiments on the solubility of illite vs. other mineral assemblages are in violation of Schreinemakers’ rules and indicate lack of equilibrium.

Predictable sequences of clay minerals as a function of temperature are best understood through the Ostwald step rule, in which clay mineral assemblages undergo reactions in response to kinetic factors that represent reaction progress rather than an approach to equilibrium. Currently used clay mineral thermometers (illite crystallinity, smectite/illite reaction, chlorite composition) are not based on equilibrium reactions. Such systems are not accurate thermometers and therefore have questionable utility.

In the Singhbhum Craton of the Indian shield, the Remal granite-gneiss preserves felsic magmatic fabrics onto which a low-temperature segregation layering has been superposed. Planar, sub-horizontal to gently dipping layers (Sign1) comprise K-feldspar megacrysts, plagioclase and quartz, with the base of each layer defined by segregations of biotite. Sign2 consists of trough cross-bedded layers composed of K-feldspar phenocrysts, plagioclase and quartz with biotite schlieren defining the base of each layer. Microstructural features such as concentrically arranged mineral inclusions in K-feldspar phenocrysts and graphic intergrowth textures testify to the magmatic origin of these fabrics, with insignificant subsequent metamorphic reconstitution. The tectonic fabric S1 has developed sub-parallel to localized greenschist-facies mylonite bands, and is defined by weakly aligned flakes of biotite. Crystallographic preferred orientations away from the mylonitized domains show a strong alignment of K-feldspar, quartz and biotite parallel to the magmatic fabric due to efficient segregation during magmatic flow. Quartz crystallographic preferred orientations within the mylonitized domains show a strong preferred orientation and dextral asymmetry. Temperature constraints from synkinematic chlorites along with estimates of deformation temperature from quartz crystallographic preferred orientations indicate that mylonitization occurred at the lower limits of quartz crystal plasticity. The results of combined thermodynamic and multiphysics modelling studies show that felsic magmas can undergo significant convective motion for a wide range of crystallinities and water contents before solidification. Additionally, segregation layering resembling a gneissosity can develop at low temperatures owing to localized mylonitization and concomitant dissolution–precipitation of biotite.

Diverse applications of chlorite thermometry have been considered for better understanding the formation process in nature. Here, an approach which combined a semi-empirical thermometer (Inoue et al., 2009) with the method of Walshe (1986) was tested to estimate the redox conditions (log fO2) and the formation temperature, using the literature data from Niger, Rouez and St Martin and new data for chlorite which coexists with pink-coloured epidote in the Noboribetsu geothermal field. The log fO2 predicted for the former data sets were compatible with those estimated by Vidal et al. (2016), suggesting that the present approach is valid for quantifying the variations in log fO2. The Noboribetsu chlorites have lower Fe/(Fe + Mn + Mg) and greater Fe3+/ΣFe ratios than those observed in adjacent propylite rocks. The peculiar mineral assemblage and chemical composition are attributed to the formation under higher fO2 conditions and possibly low Fe concentration in the alteration fluids.