Preferred orientation of clay particles in various clay mounts was measured with an X-ray texture diffractometer. Pole distributions are approximately symmetrical about the normal to the sample. Particle orientation is characterized by standard deviations ranging from 7° for thin sedimented layers, to 20° and more for dry-pressed samples, the latter showing less dependence on particle size. The orientation of the reflecting particles may span a considerable angular range affecting the intensities of diffraction peaks observed on contemporary powder diffractometers, even those with moderate axial divergence. A theory based on the diffractometer geometry was developed to calculate the effect of particle orientation. Results are presented as modified Lorentz factors for orientations deduced from experimental observations. In qualitative X-ray diffractometry of clays, preference must be given to preparation techniques which ensure a high degree of preferred orientation and therefore strong enhancement of basal reflections. In quantitative X-ray diffractometry, the main factor is the reproducibility of particle orientation, and suction and dry pressing are promising methods of sample preparation.

Six deferrated clay (<2 µm) samples from the major B horizons of pedons derived from basaltic pyroclasts (Western Cameroon) were selected. These soil clays, which differ mainly in their contents of halloysite and kaolinite, represent the types of secondary constituents characteristic of different weathering stages in soils derived from such volcanic materials under humid tropical conditions. On these clays, Ca2+-K+ exchange equilibria were performed at pH 6 and 0·01 n chloride concentration. The amounts of K+ specific sites were estimated through a two-site modelling of the exchange isotherm curves. The total and external surface areas of the clays were measured by EGME retention and by the BET treatment of the N2 adsorption isotherm, respectively. Relative halloysite content was found to be directly related to all measured surface charge properties, i.e. CEC, amount of K+ specific sites, and total and internal surface areas. It was also observed that thermal collapse of halloysitic layers reduced both the CEC and the amount of K+ specific sites. All the surface properties observed point to the presence of smectite units intimately associated with halloysite in these clays. The results further suggest that the layer charge of these smectite units differs in different samples.

Until recently, the determination of the defect structures (previously referred to incorrectly as “crystallinity”) of kaolinites has been obtained in one of two ways: (1) measurement of the Hinckley index, or (2) by comparing calculated X-ray diffraction patterns based on a model of the defect structure (including types of defects and abundances) with experimental diffraction profiles. The Hinckley method is simple and easy to perform but contains no real information about the defect structure. Calculated XRD patterns are based on real defects but these calculations are time consuming and require some skill in application. Another approach is proposed: an expert system which will accurately describe the defect structure of kaolinites based on a few measurements taken from a normal powder diffraction profile. This system has been verified for nine kaolinite samples for which the defect structure was previously determined by comparison of calculated and observed diffraction profiles. The expert system reproduced the correct defect structure for each of the samples.

Glauconites are dioctahedral, microdivided, Fe-containing silicates whose idealized structure may be described as a stacking of 2 : 1 layers with the same azimuthal orientation. The diffractometric and crystallochemical data do not coincide with this idealized structural model. For each glauconite studied the best fit between experimental and theoretical XRD patterns was obtained using the same model. It was possible to describe the stacking by an equally probable occurrence of layers rotated by multiples of 60°. The layer types alternate with a definite tendency to segregation. The model is characteristic for all glauconites, irrespective of composition, age and type of host rock. The alternation of layers rotated by 60° in the structure leads to both octahedral and prismatic coordination for the interlayer K+ cation.

Mössbauer spectra of Garfield nontronite H33a were taken at temperatures between 2·5 and 37 K under longitudinally applied magnetic fields up to 9 T. While no magnetic order was apparent in the absence of an applied field above 7 K, the application of external magnetic fields led to the induction of magnetic hyperfine splitting up to at least 19 K. Variation of the applied fields allowed determination of hyperfine fields of bulk nontronite, and indicated the bulk sample to have a Néel temperature of about 20 K. The non-ideal behaviour of this nontronite, leading to the lack of magnetic order in the absence of applied magnetic fields, is in line with the frustration of antiferromagnetic order in the octahedral sheets due to the presence of about 6% of the iron content in tetrahedral sites, although some influence of magnetic dilution cannot be excluded.

The magnetic susceptibilities of seven different nontronites in their natural oxidation states were measured between 5 and 100 K. Results revealed that the magnetic exchange interaction in all samples was antiferromagnetic, except no clear minimum occurred at the Néel temperature. Possible explanations for this phenomenon which are discussed include magnetic dilution due to isomorphous substitution, and antiferromagnetic frustration due to either non-centrosymmetric distribution of octahedral Fe3+ or tetrahedral Fe3+ substitution. A computer simulation model was developed to demonstrate the effects of these variables on long-range magnetic ordering. Magnetic dilution and tetrahedral Fe3+ content could explain the anomalous antiferromagnetic behaviour in some, but not all, samples. The non-centrosymmetric model is the only one which explains the behaviour of all samples. In this model, at least 13% of the octahedral Fe3+ would occupy trans-dihydroxide sites, with the balance in cis sites. Magnetic frustration occurs because two Fe3+ neighbours of a third Fe3+ ion are also neighbours to each other, making the simultaneous satisfaction of all antiferromagnetic exchange interactions impossible.

Six clays from volcanic ash soils at different stages of weathering differ in their relative halloysite content with respect to kaolinite and several surface properties, namely CEC, and exchange selectivity for K+. These three parameters are related to each other in that they all decrease with increasing soil weathering stage. XRD data show that the hydrated 1:1 layer-silicates in these clays combine with smectite to form interstratified H/Sm clay minerals. In these mixed-layers, the content and layer charge of smectitic units decrease as the relative halloysite content in the clay decreases. These clays thus depict a weathering sequence that is parallel to the weathering sequence of the soils from which they originate. It is also shown that the smectites in the H/Sm minerals have the distinctive composition and ESR spectrum of Fe-rich 2:1 clay minerals belonging to the beidellite-nontronite series. The information obtained explains why these clays have high CEC and distinct affinities for K+. It is hoped that this study will help to clarify the controversy concerning the CEC and related surface properties attributed to hydrated halloysite.

Calculated X-ray diffraction profiles of 10 Å-17 Å interstratified clay minerals, which make allowance for fundamental illite particles, confirm many of the inferences made in a previous analysis by Corbató & Tettenhorst (1987). However, some of these profiles do not appear to match well with patterns obtained from real interstratified clay minerals. The lack of agreement is apparently a consequence of the assumption that the interlayer spacing formed at the contact of I and S (illite and smectite) silicate layers is exclusively 17 Å. The results suggest that this interlayer spacing may be either 10 Å or 17 Å.

ESR has been used to obtain information on the octahedral or interlamellar position of Cu(II) in natural smectites from Burkina Faso (West Africa). On the basis of 060 XRD reflections and chemical data, these smectites were found to be Al-rich and dioctahedral. After both Mehra & Jackson, and De Endredy deferrification treatments, the Cu contents remained high (4500 and 22000 p.p.m., respectively). The Cu(II) ESR spectra of these deferrated smectites were compared to those of two reference smectites for which the structural position of Cu(II) was precisely known. The interlayer Cu(II) signal was obtained on a Cu-saturated Camp Berteau montmorillonite, while the octahedral Cu(II) signal was obtained on a synthetic Cu-rich smectite. For this latter reference sample, EXAFS spectroscopy provided evidence that Cu was in six-fold coordination in the octahedral sheet only, and was not exchangeable. In agreement with the experiments by Clementz, Pinnavaia and Mortland, a shift in the g⊥ ESR signal was observed when the air-dried Cu-saturated Camp Berteau montmorillonite (g⊥ = 2·05) was soaked in water for 48 h (g⊥ = 2·13). A small shift in the opposite sense was observed for the synthetic Cu smectite (g⊥ = 2·05 for the air-dried sample, g⊥ = 2·02 for the water-soaked sample). For the two natural smectites a small shift similar to that for the synthetic Cu-smectite was observed. These results indicate that up to 10% of the Cu atoms substitute for Al-Mg-Fe atoms in the octahedral sheets of the smectites studied.

A pillared smectite clay has been formed by treating Na-montmorillonite with a hydroxy-aluminium solution with an OH/Al ratio of two. Heating the pillared clay for four h at 500°C reduces the interlayer spacing, measured at ambient humidity, from ∼ 9·4 to 8·0 Å. Hydrated VO2+ (and Mg2+) ions have been introduced into the region between the pillars by immersing the pillared clay in VOSO4-MgCl2 solutions and EPR measurements show that in general both immobile and mobile VO2+ ions are present. The immobile ion is probably VO2+(OH)x(H2O)5·x where x = 2–4, and is adsorbed on the pillars with a concentration that saturates at about 1 per 8 Al13 pillars in the unfired clay, and 1 per 4·3 pillars in the fired clay. The mobile species is VO2+(H2O)5 and for both fired and unfired clay in the wet state at room temperature it has a rotational correlation time of (9 ± 1) × 10−11s rad−1.

Two-layered Fe(III)-M(II) hydroxy carbonates (M(II) = Ni or Mg) have been synthesized by induced hydrolysis using controlled air oxidation of an aqueous Fe(II)-M(II) mixture (M(II)/Fe(II) ratio >3) at a pH below which the hydroxide of the M(II) cation precipitates. The crystalline, homogeneous product can be a single phase consisting of very thin circular to hexagonal plates (0·2–0·8 µm diam.). For synthetic reevesite (Ni(II)-Fe(III) hydroxy carbonate), stable spherular aggregates are formed. The well crystallized products have an M(II)/Fe(III) ratio of about two. However, by varying the pH, the initial M(II)-Fe(III) ratio and the oxidation rate, the M(II)/Fe(III) ratio in the product and its crystallinity can be varied. The derived formulae are in reasonable agreement with the known composition of pyroaurite-type compounds. Possible synthesis pathways are discussed. The ability to control some physical and morphological features of the products indicates that the synthesis technique could prove advantageous in the preparation of certain catalyst precursors.

In order to study Al-Fe substitution, Fe-kaolinites were synthesized at 200°C from glasses (Si/(Al + Fe) = 1) with varying Fe content. XRD and IR spectroscopy showed that kaolinites with increasing Fe contents could be synthesized, and Mössbauer spectroscopy indicated that the iron is strictly ferric. Chemical analyses (STEM) of kaolinite particles determined up to almost 7% Fe2O3, which probably does not represent a structural limit. The effect of Fe3+ content is evident on growth kinetics and particle size of kaolinites, both decreasing with % Fe2O3. However, it is not possible to establish a general correlation between Fe content and percentage of structural defects. Acccording to IR spectroscopy, some synthesized Fe-rich kaolinites can have relatively good crystallinity.

Weathering processes have been studied in a podzol profile developed on sand during a postglacial period of 9400 years. Due to disintegration of rock fragments, the particle-size distribution has changed markedly in the upper part of the profile, and minerals have been transferred from sand fractions to silt and clay fractions. The most important weathering processes are the total breakdown of trioctahedral chlorite and biotite, and the transformation of dioctahedral mica (muscovite-phengite) to a regularly interstratified mineral, and further to Al-vermiculite or smectite. By using quartz as an internal standard, the annual loss due to weathering was found to be 3·2 g/m2. The release of elements due to silicate weathering was calculated from depletion curves, and the average annual release of Na + K + Mg + Ca for the postglacial period was 19 mEq/m2.

Isocitric acid oxidative decarboxylation was realized in the absence and in the presence of homoionic Na+-, Mn2+-, and Cu2+-montmorillonite. The catalytic activity of the clay depends upon the nature of the interlayer exchangeable cation. Isocitric acid is transformed into α-ketoglutaric acid under the action of the clay mineral saturated with Na+ cations which do not form a complex with the isocitrate anion. Nevertheless, the reaction rate is very much lower than in the presence of the enzymatic system (isocitrate dehydrogenase enzyme and nicotinamide adenine dinucleotide phosphate coenzyme). The reaction mechanism in the presence of clay is given showing the different steps of the transformation.

EPR spectroscopy was used to provide evidence for the existence of vanadium(IV) in two samples of rectorite. Possible sites for the vanadium are the tetrahedral and octahedral sites in the 2:1 layers and the non-exchangeable interlayer sites in the paragonite part of the structure. Consideration is given to each of these possibilities but the experimental evidence is insufficient to identify conclusively the location of the vanadium. Since substantial amounts of vanadium(IV) were present in the only well-defined rectorite samples obtainable, it is possible that the formation of rectorite is facilitated by the presence of vanadium.

An empirical method to estimate rapidly both mean total layer charge and location of charge in smectites and vermiculites is presented, involving exchange with a long-chain alkylammonium ion before and after the Greene-Kelly test. The method utilizes an empirically determined relationship between XRD basal spacings and the mean layer charges calculated from the conventional n-alkylammonium ion-exchange technique using simple linear regression models to describe the relationships for both monolayer-bilayer and bilayer-pseudotrimolecular layer transitions. The method provides a rapid estimation of these charge parameters and facilitates the proper identification of both smectites and vermiculites consistent with the AIPEA criteria for classification based on total layer charge. It is considered superior to conventional methods of differentiation based mainly on expansion and collapse behaviour.

Vermiculitization of mica is one of the main weathering processes in soils from three upland catchments receiving various levels of acid deposition. Usually this process is manifested by the presence of interstratified mica-vermiculite with the interlayer space in the vermiculite often partially filled with polymeric hydroxyaluminium species. In one peaty podzol, regularly interstratified mica-vermiculite clearly develops at the expense of mica and is the dominant mineral in the Eh horizon. It was concentrated by chemical treatments to remove organic matter, free iron oxides, and any Al species in the interlayer, and the structural formula calculated from chemical analysis confirmed the dioctahedral character of both vermiculite and mica components, and indicated that the vermiculitic weathering product was formed from a dioctahedral mica. The degree of interlayering in the interstratified phases seemed to be pH-dependent with resultant implications for soil and freshwater acidification.

A study has been carried out on the influence of acid concentration, time and temperature of reaction, liquid to solid ratio and chemical composition of the mineral on the extension and selectivity of the octahedral cations extracted during acid leaching of palygorskite. The mechanism of the extraction of Mg2+, Al3+, and Fe3+ is discussed on the basis of kinetic models fitted for a reacting-shell model, taking into account particle shape and the rate controlling step. A model is given from which kinetics parameters and the activation energy of the process are calculated. The different behaviour observed for the two palygorskites could be explained on the basis of their differences in surface areas.

The preparation of green rusts from sulphate solutions and representative Mössbauer spectra are described. As the samples oxidized readily, attention focused on the Mössbauer parameters at liquid nitrogen and helium temperatures. The spectra recorded at 77 K could be fitted satisfactorily with one ferrous iron quadrupole doublet with a separation of 2·93 mms−1 and one ferric iron quadrupole doublet with a separation of 0·45 mms−1. In some spectra a ferric iron magnetic hyperfine of strength 49·2 T was also apparent. At 4·2 K, the ferrous iron exhibited a hyperfine splitting with a field of 12·4 T whilst the ferric iron exhibited a hyperfine splitting with a field of strength 50·4 T. The ratio of ferrous to ferric ions was 2·25 ± 0·25 at 77 K and at 4·2 K, and ∼1·6 with a large variation at room temperature. The liquid helium spectra did not always give a good chi-squared fit, the main reason being attributed to relaxation. The line-width of the ferrous iron site at 77 K is slightly larger than that for iron metal and could be explained by a variation in the number of near Fe3+ neighbours at different Fe2+ sites, consistent with the assumption that the ferrous iron site is in the hydroxide sheet. The effect of different numbers of Fe2+ and Fe3+ neighbours probably contributed to the increase in line-widths at 4·2 K compared with those at 77 K. The ferrous iron doublet is marginally different to those of chloride and hydroxy-carbonate green rusts and the aluminium analogues.