Accessory green ferric clay occurs in fluvial litharenites of the Early Triassic Rewan Group. Although resembling glauconitic minerals in thin-section, electron microprobe analyses indicate that the green ferric clay is mainly ferric illite. The ferric illite may have formed in a small hypersaline lake or well drained, flood-plain soil, and its presence in the Rewan Group cannot be used to support a notion that the unit is partly marine influenced. Identification of the ferric illite by electron microprobe analysis shows that some green ferric clays, particularly those that resemble glauconitic minerals optically, may require precise quantitative elemental analysis before they can be used as a basis for environmental interpretation. Green ferric clay in the Rewan Group also includes detrital celadonite that most likely originated in a volcanic arc setting.

The distribution of clay minerals in recent sediments on the continental shelf and in the Bay of Cádiz has been analysed as part of a wider study of sedimentary exchange between the continent and the continental margin. The clay minerals included in the muddy sediments consist mainly of illite, kaolinite, smectite, randomly mixed-layered illite-smectite, chlorite, vermiculite and randomly mixed-layered illite-chlorite. Distribution of the minerals indicates that inheritance from the adjacent continental areas is the most important process involved.

The relations between the clay minerals were established by Q-mode and R-mode factor analysis. The main clay mineral associations are: (1) illite-kaolinite, (2) smectite-random mixed-layered illitesmectite, and (3) kaolinite-random mixed-layered illite-smectite-vermiculite. The main sources of sediment supply to the continental shelf are the Guadalquivir and Guadiana rivers. Sediment from these rivers and from the rivers flowing into the Bay of Cádiz (mainly the Guadalete) is transported to the part of the shelf outside the bay by ebb-tide currents. The path of this outflow to the continental shelf, where the sediment is partially deposited, is influenced by the morphology of the coast and by hydrodynamic processes.

A mineralogical study and multivariant geochemical analysis have been carried out to determine the sedimentation processes and the influence of surrounding rocks on sediments. Analyses were made on the clay fraction of sediments and on bulk samples of rocks from nearby reliefs in the Sorbas Basin (SE Spain). Source rocks were identified by statistical analysis of trace elements and by comparison of the crystallographic parameters of phyllosilicates in sediments and supposed sources. Palaeoenvironmental conditions are recorded by specific elements in the clay minerals. The study of the B and Mg content and of the different types of smectites and palygorskite found in distinct areas has permitted a description of these conditions for this Basin. In this way, different types of deposition environment, showing varying degrees of confinement and influence from marine or continental water, have been distinguished.

Transformations of clay minerals were studied in a sequence of podzols (Entic Cryorthods) of increasing age (5500-10000 years BP) developed from till materials in central Finland. The age of the soils was estimated according to altitude above the present sea level, with reference to the age of ancient shorelines and the rate of land uplift. Clay minerals in fine (<0.1 µm), coarse (0.1-2 µm) clay and fine silt (2-5 µm) fractions were studied by X-ray diffraction. The layer charge of expandable minerals was estimated on the basis of their re-expansion with ethylene glycol following K-saturation and heating to 110°C Pedogenic transformations of clay minerals occur essentially in the E horizons of the podzols. Smectites and illite-smectite mixed-layers are the dominant clay minerals in the E horizons of soils older than 6500 years. The number of mixed-layers decreases with evolution of the soil, leaving a nearly pure smectite phase in the oldest soil (10000 years). The smectites are dioctahedral with a tetrahedral charge. Groups of low-charge and highcharge interlayers are both present in the samples, but low-charge smectite layers were found only in the two older soils (9500-10000 years). This suggests alteration with time of the high-charge smectites.

The clay mineralogy and surface charge characteristics of four basaltic soils from Western Samoa have been studied. The soils contained subordinate to dominant amounts of poorly ordered allophanic material in addition to varying amounts of crystalline free iron oxide minerals. Infrared studies revealed the presence of trace to subordinate amounts of kaolin minerals in all the soils. The surface charge-pH curves followed a constant potential model indicating the presence of substantial amounts of pH-dependent charge. Some negative charge was present, however, at pH values as low as 3.0 and small quantities of positive charge were detected at pH values as high as 9. Values for PZC ranged from 2.2 to 3.9 and these were generally higher than the pHo determined by the ΔpH method.

For the purpose of clay examination, a new one-step method for the removal of organic matter was developed using sodium peroxodisulphate combined with different buffers such as sodium hydrogen carbonate, disodium hydrogen phosphate and disodium tetraborate. From an early Cretaceous black shale from the Apennines with a high organic carbon content, the <2 µm clay fraction was separated and contained 10.9 wt% organic carbon. To prevent decomposition of the clay layers, the period of oxidation was short (25-60 min) and the pH was kept between 5 and 9.5. Up to 98% of carbon was removed by this method.

A range of allophane samples (atomic AI/Si bulk ratios 1.1-1.9) from Silica Springs, New Zealand, has been characterized by X-ray photoelectron spectroscopy (XPS). Binding energies of Si 2s, Si 2p, Al 2p, O 1s, C 1s, and N 1s electrons, together with the kinetic energies of Al KL23L23 Auger electrons, at or near the surface of allophane aggregates, have been derived. The values for Al, Si and O electrons are similar to those for kaolinite but also to those for some framework silicates (feldspars) having 4-coordinate Al. Values for N electrons suggest that N occurs in organic structures. Comparison of XPS and bulk Al/Si ratios shows an enrichment of Al at or near the surface of allophane aggregates. The same is true for C and N. Extraction with citrate-dithionite-bicarbonate (CDB) reagent leaves the surfaces depleted in Al. The CDB extracts have higher Al/Si ratios than the bulk allophanes. Similarly, CDB treatment reduces the degree of surface enrichment of C and N. Small increases in the binding energies of Si electrons following CDB treatment suggest partial dissolution of the bulk structure though a concomitant removal of a separate phase or species cannot be ruled out. The results may be accounted for in terms of the structure previously suggested for the primary spherules of Silica Springs allophane (Childs et al., 1990) though the composition of the spherules at or near the surface of the allophane aggregates is different from those of the bulk.

Dehydration and rehydration of a phlogopitic vermiculite saturated with either Sr2+ or Ba2+ were studied by thermoanalysis (TG, DSC, DTA) and by in situ X-ray powder diffractometry. A detailed analysis of the relationship between temperature and state of hydration of the vermiculites close to equilibrium at atmospheric pressure was rendered possible by use of low heating rates in combination with a fast recording position-sensitive detector. For Sr-vermiculite, the results reveal the existence of a number of distinct states of hydration characterized by basal spacings of 1.54, 1.51, 1.49, 1.23, 1.21 and 0.96 nm. The Ba-vermiculite failed to develop a two-layer hydrate even in contact with water. Its one-layer hydrate (1.24 nm) was reversibly transformed into an incompletely dehydrated ordered ‘mica-structure’ (1.00-0.99 urn), before an irreversibly dehydrated structure was attained at temperatures exceeding 700°C The formation of regular 1:1 interstratification was confirmed for all transformations characterized by a change in the number of water layers. The concluding compilation summarizes basal spacings, stacking orders, layer translations and specific water contents for all discrete states of hydration observed in our experiments at saturation of vermiculite with different alkaline earth cations.

Changes in the molecular structure of a highly ordered kaolinite, intercalated with urea and potassium acetate, have been studied using Raman microscopy. A new Raman band, attributed to the inner surface hydroxyl groups strongly hydrogen bound to the acetate, is observed at 3605 cm-1 for the potassium acetate intercalate with the consequential loss of intensity in the bands at 3652, 3670, 3684 and 3693 cm-1. Remarkable changes in intensity of the Raman spectral bands of the low-frequency region of the kaolinite occurred upon intercalation. In particular, the 144 and 935 cm-1 bands increased by an order of magnitude and were found to be polarized. These spectroscopic changes provide evidence for the inner surface hydroxyl group-acetate bond being at an angle approaching 90° to the 001 face. Decreases in intensity of the bands at 243, 271 and 336 cm-1 were observed. The urea intercalate shows additional Raman bands at 3387, 3408 and 3500 cm-1 which are attributed to N-H vibrations after formation of the urea-kaolinite complex. Changes in the spectra of the inserting molecules were also observed.

Ni-Fe pyroaurite-type hydroxychlorides were prepared by aerial oxidation of Ni(II)-Fe(II) hydroxides precipitated in aqueous solution with various P = Fe/Ni ratios. When P≥ 1/3, Ni(II)-Fe(II)-Fe(III) hydroxychlorides characterized by a specific Fe(IIl)/[Fe(II)+Ni(II)] ratio of 1/3, corresponding to the idealized formula of NiII3-xFeIIxFeIII(OH)8Cl.nH2O (with 0 ≤ x ≤ 3), were obtained at the end of the first stage of oxidation, In a second reaction stage, these hydroxychlorides oxidize with deprotonation of hydroxyl ions into O2- ions, i.e. the remaining FeII(OH)2 groups are transformed into FeIIIOOH groups. Along with the Ni(II)-Fe(III) hydroxychloride which contains a part of the FeIIIOOH groups a second phase is obtained. It is an amorphous Fe(III) or Ni(II)-Fe(III) oxyhydroxide when 1/3<P≤3/2, and a ferric oxyhydroxide identified as γ-FeOOH (lepidocrocite) when P>3/2. On the other side of the domain, when P<1/3, the Fe(III)/[Fe(II)+Ni(II)] ratio cannot reach the specific value of 1/3; this gives rise to a pyroaurite-type Ni(II)-Fe(III) hydroxychloride with a lower chloride content, that is with an average composition of NiII3+yFeIII1-y(OH)8Cl1-y.nH2O where y = {[4/(1+P)] - 3}, down to minimum Fe(III) and Cl contents corresponding to y = 1/3 (P = 1/5). The in situ mechanisms of oxidation of Ni(II)-Fe(II) hydroxides into Ni(H)-Fe(II)-Fe(III) hydroxychlorides are discussed.

Heating experiments were carried out with ferrihydrite, in the presence of organics, to gain more insight into the intermediate formation of a ferrimagnetic cubic phase often observed in soils as the result of fire. Without an organic reductant, no intermediate cubic phase was formed when ferrihydrite was transformed into hematite, regardless of whether the ferrihydrite DTA curve showed one or two exothermic peaks. Addition of glucose or charcoal as reductant caused the initial formation of a cubic magnetite and/or magnetite/maghemite phase in samples heated above 300°C in both air and N2 atmospheres. Ferrihydrite reduction, as determined by increasing cubic unit-cell edge lengths between 0.832 and 0.840 nm, increased with reductant concentration and heating time, providing excess reductant remained. Following consumption of the reductant, FeII may be partially or completely reoxidized, forming maghemite and then hematite. The occurrence of maghemite, but not of magnetite, in soils where forest fires have occurred suggests that sufficient oxygen was available, while temperatures remained elevated, to oxidize magnetite to maghemite.

Hydrogen-forms of <2 µm fractions of six bentonites of various Fe contents were prepared by H+→OH-→H+ ion exchange using resins. Potentiometric titration curves revealed that the number of strong acid sites varied and accounted for 60-95% of the total acidity in the freshly prepared H-forms. The number of strong acid sites decreased and that of the weak acid sites increased on ageing. The process of autotransformation in aqueous dispersion at 90~ was completed within four days. Layer-charge distributions of all samples were inhomogeneous with layer charges from 0.25-0.39 Eq/unit O10(OH)2. Oxalate pretreatment of the samples resulted in changes in the layer-charge distribution due to the removal of readily soluble phases which may have blocked exchange sites. After autotransformation, the alkylammonium exchange method revealed inhomogeneous charge density distributions; the fraction of layers of the highest charge decreased. Comparison of total CEC obtained from potentiometric curves and interlamellar CEC calculated from the mean layer charge confirmed attack of protons from particle edges. However, for several samples the structural attack may also occur from the interlayer space. Autotransformation of the Hsmectites decreased the mean layer charge. Protons probably attack the Mg(O,OH)6 octahedra preferentially during the autotransformation.

A natural saponite was acid activated at room temperature or 90°C with different acid/clay ratios and the products were characterized by powder X-ray diffraction, infrared spectroscopy and thermogravimetry. The leaching of Mg from the octahedral sheets is enhanced by an increase in the acid/clay ratio and by an increase in temperature of activation. Textural properties are reported, and it appears that they are strongly correlated to the presence of a noncrystalline silica phase which is formed during the acid activation process. The desorption of cyclohexylamine indicates that for samples activated at 90°C the number of acid sites in the acidactivated saponites decreases following severe acid treatment. Infrared spectroscopy of adsorbed pyridine on samples after calcination at 500°C suggests that acid activation at 90°C produces a single type of Bronsted site but two types of Lewis sites whereas activation at room temperature results in only one type of Lewis site in addition to a Brønsted site. The two Lewis sites are suggested to originate from residual Al in the clay structure and to AI exsolved from the layers during activation. The dehydration of pentan-1-ol has been used as a further probe to measure acidity by monitoring the degree of conversion and selectivity for the different samples.

Montmorillonites ion-exchanged with Li+, Na+, K+ Ca2+, Mg2+ and Ni2+ and acidic clay were used as catalysts for the polymerization of a cyclic siloxane monomer, 2,4,6,8- tetramethylcyclotetrasiloxane. Montmorillonites with Ni2+ and Mg2+ in the interlayer, and acidic clay exhibit a greater ability for siloxane polymerization in both yield and mean molecular weight of products than those containing Li+, Na+ and K+. The difference in catalytic ability of the ionexchanged montmorillonites is caused by the number of Brmasted acid sites due to the polarization of H2O. This was confirmed experimentally by FTIR analysis of pyridine-treated samples. Therefore, it may be possible to design a catalyst for controlling siloxane polymerization (i.e. mean molecular weight of product) by changing the number of Bronsted acid sites through exchange of the interlayer cations of montmorillonites.

The adsorption of the cationic dye acridine orange (AO) by Na-saponite and the colloidal properties of the aqueous suspension were investigated by visible spectroscopy and XRD. The organic cation is adsorbed by the mechanism of cation exchange. When small amounts of the dye are adsorbed, the system contains small tactoids and is peptized. At this stage the dye penetrates into the interlayer space and most of it undergoes metaehromasy due to interactions between the aromatic entity and the oxygen plane of the clay. When greater amounts of AO are adsorbed, the clay platelets flocculate to form book-house floes which, with excess AO, are transformed into card-house floes. At this stage metachromasy results from the aggregation of the dye in the interparticle space of the floes, in addition to the π interactions with the oxygen plane. In excess AO, the clay is gradually peptized. At this stage the dispersed clay platelets form small tactoids with monomeric AO in the interlayer space and at the same time adsorb dimerie and polymeric AO cationic species at the solidliquid interface.

A series of V-doped titania-pillared clay catalysts, characterized by ICP-AES chemical analysis, X-ray diffraction, BET surface area measurement, and ESR spectroscopy, have been tested in the selective catalytic reduction of NO by NH3. An ESR analysis shows that V dopant is anchored to the titania pillars. Vanadyl species with differing degrees of in-plane V-O π-covalent bonding are produced depending on the method of sample preparation. Polymeric V species appear as the V content is increased. Catalytic performance of these systems depends on the method of preparation and on the V content. The best catalyst, converting 90-100% NO in the temperature range 523-623 K, is obtained by exchange of pillared montmorillonite with vanadyl ions, at an extent of exchange below the level where significant amounts of polymeric V species appear. The co-pillared catalyst, containing vanadyl centres characterized by a higher degree of in-plane ncovalent bonding (according to ESR), is less selective than the exchanged samples.