The structure of an illite-smectite (I-S) sample (HSI) from the Tertiary–Cretaceous boundary layer at Stevns, Denmark, i.e. the so-called Fish Clay, is investigated in order to discuss the origin of this clay and its similarity to clays from Maastrichtian-Danian chalk of Denmark and the North Sea. The phase compositions and layer sequences have been determined by X-ray diffraction. The structural formulae are determined from chemical analysis and solid-state 27Al magic-angle spinning (MAS) nuclear magnetic resonance (NMR) spectroscopy. The octahedral vacancy pattern has been determined by thermal analysis and the particle shape by atomic force microscopy. The HSI sample I-S consists of two phases, a high-smectitic (HS) phase (70%) having 95% smectite and 5% illite (leucophyllite) layers, and a low-smectitic (LS) phase (30%) having 50% smectite and 50% illite (leucophyllite) layers. Both phases have trans-vacant dioctahedral sheets and contain only traces of IVAl, the charge of the illite (leucophyllite) layers being provided predominantly by octahedral Mg for Al substitution. The structure of the tetrahedral and octahedral sheets for the HSI are compared to the structure of an I-S sample (GRI) from the Maastrichtian chalk at Stevns below the Tertiary-Cretaceous boundary and to the standard smectite SAz-1 from Arizona. For all three samples the 27Al MAS NMR spectra show the presence of two resolved IVAl resonances, which indicate the presence of two different IVAl sites. The NMR, infrared spectroscopy and chemical data show that both HSI and GRI have highly ordered Mg for Al substitution in the octahedral sheet, each Al having one Mg and two Al neighbours, whereas substitution of Mg for Al in SAz-1 is random. For the HSI sample, both the HS and LS phases are probably formed from volcanic ash. The structural similarity of the phases in HSI and GRI shows that GRI formed from a material similar to HSI by illitization of the HS phase and chloritization of the LS phase.

Kaolinite and synthetic γ-Al(OH)3 (gibbsite or hydrargillite) were reacted with phenylphosphonic, phenylphosphinic and 2-nitrophenol-4-arsonic acids. The products were studied by powder X-ray diffraction, transmission electron microscopy/selected area electron diffraction/ energy dispersive X-ray/Fourier transform infrared and simultaneous thermogravimetric/differential thermal analysis. The acids were not intercalated but, instead, easily destroyed the structure of the minerals. Lamellar Al phenylphosphonate and aluminium phenylphosphinate and phenylarsonate with polymeric linear-chain structures were formed from kaolinite. The reaction between gibbsite and the same acids yielded almost identical products. No evidence of formation of grafted kaolinite derivatives after the reaction with phenylphosphonic acid was found.

The mineralogy and geochemistry of the clay fraction of Victoria Land Basin (Ross Sea, Antarctica) sediments was investigated, to determine the origin of clay minerals and the features of authigenic smectite. The investigated core (CRP-3) is ~800 m long, mostly of Oligocene age. The clay fraction of the upper sequence consists of mica, chlorite and detrital smectite, while that of the central and lower part is largely made up of authigenic smectite. Authigenic smectites are ditrioctahedral, with a composition close to saponite, while detrital smectites such as Al-Fe beidellites are dioctahedral. Authigenic smectites have no illite mixed layers, show a higher degree of crystallization, higher MgO, Fe2O3, V, Cr, Co, Ni and Sc contents and lower SiO2, Al2O3, K2O, TiO2, Ba, Rb and Zr contents with respect to detrital clay minerals, and a clear depletion of LREE with respect to HREE. Authigenic smectite formed from the alteration of volcanic materials and clay minerals.

Initial stages of the gel-to-kaolinite transformation were studied using HRTEM. Spherical aggregates of kaolinite crystals were produced during the synthesis of kaolinite by hydrothermal treatment of Si-Al amorphous gels. Prior to sphere formation, gels transform into pseudospherical domains that have a Si/Al ratio of one and display no SAED pattern. The spherical particles consist of radially arranged sectors of stacks of planar crystallites. Crystals display nonbasal spacings of 4.5, 4.2 and 3.8 Å and a basal spacing of 7.1 Å , the c* axis following the radius of the sphere.

Interpretation of the 3D nanostructure of the spheres is difficult. The c* axis exhibits a radial disposition, but the relative orientation of the a* and b* axes in neighbouring crystallites may produce bent layers or incoherent contacts. In addition, curved layers with a d spacing of 7.4 Å may be attributed to halloysite layers collapsed under microscopy conditions. The disappearance of the spheres during the hydrothermal treatment is probably due to preferential dissolution of either highstress areas near bent layers or non-crystalline material filling crystal boundaries. Dissolution leads to sphere disaggregation and allows the component columnar crystals to continue to grow. Observations of the gel matrix suggest that under our experimental conditions kaolinite crystallizes via an in situ transformation of the gel.

2-line ferrihydrite stored in water at ambient temperatures from 4 to 25°C and at ten different pH values between 2.5 and 12 for up to 10–12 y transformed to both goethite and hematite at all temperatures and pH values except at pH 12 where only goethite was formed. The rate and degree of transformation (20–100%) increased with increasing pH and temperature. The hematite/ (hematite+goethite) ratio varied between 0 and ~0.8, increased with increasing temperature and showed a strong maximum at pH 7–8 which increased from 0.1–0.2 at 4°C to 0.7–0.8 at 25°C. The maximum coincides with the zero point of charge of ferrihydrite where its solubility and, thus, its via-solution transformation rate to goethite are minimal. We assume, therefore, that in this pH-range the (slower) via-solution transformation to hematite can more efficiently compete with that to goethite.

Pyrophyllite from a Triassic sedimentary formation from the Coastal Range of Chile has been investigated by transmission/analytical electron microscopy (TEM/AEM). The mineral assemblage includes pyrophyllite, muscovite, paragonite, a kaolin mineral, boehmite, rutile and hematite. The textures indicate that the protolith was a volcanoclastic rock. Petrographic evidence, chemistry, and the mineral assemblage suggest the intense leaching of the parent rock by a weathering process, before the metamorphic episode, to create the protolith for the pyrophyllite. Pyrophyllite always grows from the kaolin mineral, and both phases show close orientation relationships. The presence of parallel intergrowths of pyrophyllite and muscovite indicate that muscovite also grew from the kaolin mineral. Nevertheless, the composition of muscovite suggests that this phase must also form from another precursor, probably Al smectite. The AEM data and textural relationships between pyrophyllite and muscovite reveal the presence of two generations of muscovite and suggest that Na-rich muscovite recrystallized into a Na-free muscovite and paragonite.

Fe-rich smectites from lateritic weathering profiles have previously been studied by XRD, ICP-AES, SEM-EDX and TEM-EDX analyses (Gaudin et al., 2004). These smectites exhibit intermediate chemistries between five end-members: Al-Fe beidellites, Al-Fe montmorillonites and Mg+Ni-saponite. The spectroscopic study by FTIR and XAS of these smectites reveals that: (1) tetrahedral Fe3+ is near or below the detection limit (0.05 cation for 4Si); (2) the large chemical variability is due to substitution of the three major cations (Fe, Al, Mg) within adjacent octahedra; (3) Ni is not concentrated in another clay phase such as Ni-kerolite and is located in the octahedral sheets of smectite; (4) octahedral cations are not randomly distributed but ordered in separated Fe, Al, Mg, Ni clusters; (5) the Mg-Ni saponite end-member actually appears as small trioctahedral clusters of Mg and Ni distributed within the dioctahedral smectite.

This paper describes the swelling properties of two highly compacted clays, natural, untreated Wyoming montmorillonite (MX80) and Fourges smectite (FoCa7), saturated with Na and Ca, respectively.

The initially compacted samples were hydrated by subjecting them to different suction pressures in a range between 100 MPa and 1 kPa. For each equilibrium state, the volume change (swelling) and water content (hydration) were measured. The samples were then studied by X-ray diffraction using a transmission device to determine interlayer distance and particle size, in order to clarify both the swelling and hydration mechanisms. The distances between clay layers ranged between 10 and 21.6 Å , i.e. corresponding to between 0 and 4 water layers. Upon hydration, the particle size decreased from 350 and 100 clay layers per particle to 10 layers per particle when the suction pressure decreased from 100 MPa to 1 kPa for MX80 and FoCa7, respectively. The first swelling stage is described as being an insertion of water molecules between the layers. Then a division of the initial particles into particles of smaller size with increasingly large inter-particle distances was observed. Observations by transmission electronic microscopy confirmed these results.

Palaeozoic to early Mesozoic rocks of both autochthonous and allochthonous units of the Eastern Tauride Belt have different textural and mineralogical features related to their varying deposition regimes and thermal histories. The Devonian and Carboniferous formations in the allocthonous units have anchizonal to epizonal grades of metamorphism, whereas those of the autochthon experienced only early diagenetic conditions. A well-developed penetrative slaty cleavage is also found in Devonian formations of the allocthonous units, which is not observed in equivalent rocks of the autochthon. Phyllosilicates of the uncleaved diagenetic assemblages consist mainly of illite, chlorite, kaolinite, mixed-layer illite-smectite, chlorite-vermiculite and chloritesmectite, whereas in addition to these, metamorphic allochthonous units contain some index minerals such as dickite, paragonite, Na and K mica, pyrophyllite, stilpnomelane and chloritoid. These latter assemblages show a higher degree of textural and mineralogical maturation and lower b cell dimensions than the autochthonous rocks, suggesting that they originated in a relatively high heat flow basin such as an extensional setting.

Previous attempts to use illite as a geothermometer have failed: no general relation between the mineral's chemical composition and temperature of crystallization has been found. Here, chemical compositions of 27 illite samples from five different geothermal fields (the data on four of which were drawn from the literature) were compared with their crystallization temperatures. As previously reported by Cathelineau (1988), the K content was found to be the only variable yielding a suitable correlation, but only when applied to one geothermal field; when various geothermal systems were considered, the correlation weakened considerably. Introduction of a correction algorithm to the K content of the illite has made it possible to draw a single line to fit the data from all the studied samples, yielding a good correlation coefficient (r = 0.84).