The fraction <20 μm of a bentonitic material was selected to study the transformation of smectite into illite under hydrothermal conditions. Powder XRD showed that the studied material was composed of 97% dioctahedral smectite and 3% of quartz, cristobalite and plagioclase. The XRD analysis of the oriented, glycolated sample showed that the smectitic phase was actually a randomly interstratified illite/smectite with 11% illite. The chemical analysis of the sample yielded the structural formula: K0.25Na0.25Ca0.12Mg0.13(Al2.76Fe0.33Mg1.03)(Si7.66Al0.34)O20(OH)4. The reaction conditions of the hydrothermal experiments were: KCl concentrations 0.025, 0.05, 0.1, 0.3, 0.5, 1 M; temperature 60, 120, 175, 200°C; run time 1, 5, 15, 30, 90, 180 days; solid: solution ratio 1:5. XRD of run products showed an important transformation of smectite into illite, in contrast with all other techniques (DTA, IR, NMR), which did not detect any transformation of the original material. This suggests that illite quantification in I/S by means of XRD patterns after hydrothermal experiences can be much affected by other variables. The original material and the run products were analysed by means of DTA. The starting material displayed two dehydroxylation peaks, one at 560°C corresponding to illite, and the other at 650°C corresponding to smectite. Quantification of the two dehydroxylation peaks also yielded a content of 11% illite. The run products displayed DTA diagrams in which the smectitic peak remained unaltered, but the illitic one was reduced or eliminated by the presence of exchangeable K in the interlayer. When exchangeable K was removed the illitic dehydroxylation peak appeared again. The coherence of XRD and DTA in quantifying the proportion of illite layers and the effect of exchangeable K on dehydroxylation of illite layers in mixed-layer illite/smectite seem to be positive proofs of the existence of polar layers.

Offscraped strata within the toe of Nankai accretionary prism display an overall facies pattern of thickening and coarsening upward. Detrital clay minerals within the Quaternary trench-wedge facies are dominated by illite; chlorite is the second-most abundant clay mineral, followed by smectite. Relative mineral percentages change only modestly with depth. The hemipelagic clay-mineral population is virtually identical to clays washed from turbidite matrix, and different size fractions (<2 μm and 2–6 μm) show nominal amounts of mineral partitioning. Smectite content increases beneath the trench-wedge deposits, where the upper subunit of the Shikoku Basin stratigraphy (late Pliocene and early Pleistocene) includes abundance of volcanic ash. Syneruptive, subaerial chemical weathering of volcanic source rocks, together with in situ alteration of disseminated glass shards, caused the increase in smectite. Smectite begins a monotonic transformation to illite/smectite (I/S) mixed-layer clay at ~555 mbsf and an estimated temperature of ~65 °C. Ordered (R = 1) I/S interlayering first appears at ~1220 mbsf (<2 μm size fraction) and ~1100 mbsf (<0.2 μm size fraction). The illitization gradient coincides with a reduction in pore-water chlorinity, but depth-related changes in bulk mudstone geochemistry (K2O, Rb) are subtle. The absolute abundances of discrete smectite and I/S appear to be insufficient to account for the magnitude of pore-water dilution via in situ dehydration reactions. Instead, pore water probably was transported to Site 808, either from sources located deeper in the accretionary prism, where bulk mudstone porosities are lower, or from strike-parallel sources where mudstones originally deposited in the Shikoku Basin might contain higher percentages of smectite.

The mineralogy and geochemistry of shales reflect the composition of the initially deposited precursor mud, subsequently modified by diagenetic processes. To see if significant geochemical differences exist between shales that mainly owe their present-day composition to either deposition or diagenesis, we compare the published mineralogical, bulk and clay-fraction geochemical, and clay-fraction O-isotopic compositions of 2 shales. One shale is from the Western Canada Sedimentary Basin (WCSB), and its composition mainly reflects primary (depositional) chemical and mineralogical variations (smectitic to illitic illite/smectite) within this unit. The other shale is from the United States Gulf Coast (USGC), and its composition mainly reflects mixed-layer illite/smectite (I/S) diagenesis of deposited smectitic clay material. The chemical and mineralogical trends of WCSB and USGC shales, including one of increasing illite content in I/S with depth or maturity, are essentially indistinguishable, in both bulk shale and clay fraction, despite the contrasting genetic interpretations for the origin of the contained I/S. Thus, similar mineralogical and chemical trends with depth or temperature can result either from inherited depositional compositional heterogeneity of the sediment, from burial metamorphism of shale or a combination of both. Interestingly, the O-isotopic compositions of the clay fractions from the WCSB and USGC are significantly different, a fact that reflects original clay formation from source material and water of quite different isotopic compositions. The discrimination between depositional and diagenetic contributions to shale composition continues to pose challenges, but a combination of bentonite, illite polytype, clay isotopic and trace and rare earth elemental analyses together with illite age analysis holds promise for future work.

The Middle Proterozoic (1.1 Ga) Nonesuch Formation, host of the stratiform copper deposit at White Pine, Michigan, consists of 200 m of principally dark grey clastic sediments which contain detritus obtained dominantly from underlying mafic to intermediate volcanic rocks. Clay minerals from samples collected from the mine area and drill holes up to 100 km away have been studied using SEM, EMPA, TEM and AEM. Two morphologies of phyllosilicates, both including white mica and chlorite, occur in the ‘lower’ Nonesuch Formation: (1) detrital-shaped and (2) matrix. Detrital-shaped phyllosilicate grains are up to 450 microns long with long axes parallel to bedding. Matrix phyllosilicates occur as packets typically <200 Å thick and as pore-filling cement.

TEM images of detrital-shaped chlorite generally display 14-Å periodicity, although 24-Å corrensite- like units occur locally. Most detrital-shaped chlorite from the mine area samples have a relatively restricted range of Fe/(Mg+Fe) ratios from 0.52 to 0.58, but the Fe/(Fe+Mg) ratios of detrital-shaped chlorite outside the mine area range from 0.27 to 0.64. TiO2 crystals occur within and surrounding the detrital-shaped chlorite. Matrix chlorite has Fe/(Fe+Mg) ratios of 0.47 to 0.63, indicating that it is relatively homogeneous and enriched in Fe compared to detrital-shaped chlorite.

Detrital-shaped white mica occurs as a 2Mi polytype and generally has a phengitic composition. Matrix illite-rich I/S occurs as a 1Md polytype, is K and Al deficient relative to end-member muscovite and contains significant Fe and Mg.

The data are consistent with homogenization of detrital-shaped chlorite in the White Pine mine area by hydrothermal fluids during copper mineralization. The TiO2 crystals and corrensite-like units in detrital- shaped chlorite imply that it is at least in part derived from alteration of biotite. The presence of immature 1Md illite-rich I/S and a one layer chlorite polytype with stacking disorder suggests that the matrix clays are in their original, post-smectite state of formation as consistent with an authigenic origin during early burial diagenesis; i.e., they have not undergone subsequent transformation even though sedimentation and ore deposition occurred prior to 1 Ga.

Ultrathin sections of reference 2:1 layer silicates treated with octadecylammonium cations were examined using high-resolution transmission electron microscopy (HRTEM) to establish the layer structure. Hitherto, few HRTEM ultrathin-section data existed on the expansion behavior of smectite-group minerals with different interlayer-charge values. Without such information, the expansion behavior of both low-charge and high-charge smectite minerals cannot be characterized and the structures observed in HRTEM images of clay-mineral mixtures cannot be interpreted reliably. Reference smectite-group minerals (Upton, Wyoming low-charge montmorillonite; Otay, California high-charge montmorillonite; a synthetic fluorohectorite; and a Jeanne d’Arc Basin offshore Newfoundland clay sample) with a range of layer charge values were examined. To prevent possible intrusion of epoxy resin into interlayers during embedding, the clay samples were first embedded in epoxy, sectioned with an ultra microtome, and then treated with octadecylammonium cations before examination using HRTEM. Lattice-fringe images showed that lower-charge (<0.38 eq/O10(OH)2) 2:1 layers had 13–14 Å spacings, whereas higher-charge (>0.38 eq/O10(OH)2) 2:1 layers had 21 and 45 Å spacings. These differently expanded silicate layers can occur within the same crystal and an alternation of these layer types can generate rectorite-like structures. For comparison, clay samples were also treated with octadecylammonium before epoxy embedding and sectioning and then examined with HRTEM. These samples mostly had highly expanded interlayers due to epoxy intrusion in the interlayer space. The reference clay minerals embedded in epoxy resin, sectioned, and treated with octadecylammonium cations were used to characterize smectite-group minerals in a natural clay sample from the Jeanne d’Arc Basin, Eastern Canada. Smectite-group minerals in this sample revealed similar structures in lattice-fringe images to those observed in the pure reference clay samples. Rectorite-like structures observed in lattice-fringe images were in fact smectite crystals with short, alternating sequences of low-charge and high-charge smectite layers rather than illite-smectite (I-S) phases with expanded smectite layers and non-expanded 10 Å illite layers.