Gabbroic saprolite at Mont Mégantic, Quebec, was studied in detail mineralogically to gain a better understanding of the origin of ferruginous smectite reported previously from these rocks. The parent rock is a ferrogabbro composed of plagioclase, augite, calc-alkalic amphiboles, biotite, olivine, magnetite, ilmenite, and apatite. Extensive weathering has decomposed most of the mafic minerals and magnetite to goethite, lepidocrocite, and iron-rich clay minerals, which occur in numerous microcracks distributed irregularly in the outer shells of boulders and in cracks and fissures in the bedrock. Some of the felsic minerals have altered to kaolinite. The secondary minerals and the more resistant primary minerals, such as plagioclase, ilmenite, and apatite, have subsequently moved to the lower part of the saprolite. The major ferruginous clay minerals present are smectite and vermiculite, which are compositionally similar, except that the smectite is slightly richer in SiO2 and MgO and poorer in Fe2O3 than the vermiculite.

To establish a structural formula for the ferruginous smectite, the oxidation state of Fe in a sample treated with dithionite-citrate-bicarbonate was examined by Mössbauer spectroscopy. All structural iron was found to be ferric. The calculated structural formula of a Na-saturated sample is:

$$N{a_{0.61}}\left( {M{g_{1.39}} + Fe_{0.85}^{3 + }A{l_{0.17}}M{n_{0.03}}} \right)\left( {S{i_{3.49}}A{l_{0.51}}} \right){O_{10}}{\left( {OH} \right)_2},$$

Aqueous gels of unaltered (oxidized) and chemically reduced ferruginous smectite (SWa-1 from the Source Clays Repository of The Clay Minerals Society) were characterized by transmission electron microscopy, electron diffraction, and energy-dispersive X-ray fluorescence to establish details regarding their texture, inter-layer and inter-particle arrangements, and chemical composition. Micrographs revealed that the reduction of structural Fe(III) to Fe(II) caused a consolidation of smectite particles from an extensive network of small crystals (1–6 layers thick) to distinct particles of limited size in the a-b direction and about 20–40 layers thick. The interlayer distances in the reduced sample appeared to be more uniform than in the oxidized sample, but both exhibited spacings of about 12.6 A. Chemical analysis showed no qualitative differences as a result of oxidation state. Electron diffraction patterns displayed marked differences. The pattern of the oxidized sample consisted of homogeneous rings, indicating that the stacking order in the a-b plane was turbostratic or disordered, whereas the reduced pattern exhibited much more order as evidenced by distinct spots amid low-intensity rings, suggesting that inter-layer attractive forces were stronger if Fe(II) was present in the clay crystal.

The clay mineralogy of an oxisol-saprolite overlying serpentinite and underlying basalt was studied with different techniques to evaluate the clay mineral transformation that occurred and to understand the origin of Fe3+-rich smectite. The saprolite zone of the oxisol, up to 30 m thick, contains smectites of the montmorillonite-beidellite and montmorillonite-nontronite series, as well as illite, chlorite, talc, and goethite or amorphous oxyhydrates. Illite is mainly concentrated in the upper 50 cm thick zone underlying the basalt layer and chlorite-content increases toward altered serpentinite at the base. Minor amounts of nontronite formed mostly toward westward exposures where the hot contact layer between serpentinite and basalt is only 20 cm thick. Greene-Kelly Li-tests revealed that all samples contain montmorillonite, but one sample shows the presence of a minor amount of beidellite.

Parent rocks are a mixture of mainly mica schist (the source of beidellite), and minor serpentinite in different percentages and laterally distributed. These rocks were intensely weathered under humid climatic conditions. Silica was concentrated as amorphous transparent (pure silica) cobbles and milky quartz pebbles, and originated from geothermal solutions rising through the Ovaclk thrust fault. The Mg partly formed chlorite. Ferrian smectites in serpentinites were derived obviously from the Mg-rich minerals but Mg is lost much more rapidly than Si during the formation of the clay deposit. The structural formula of the most Fe-rich smectite samples from the study area is (Si6.60−7.10Al1.40−0.90)(Al2.54−1.22Mg0.32−0.92Fe3+1.18−1.68−Ti0.06−0.04)(Ca0.16−0.10Na0.02K0.02−0.12)O20(OH)4. This composition is within the range recorded for the ferrian montmorillonite-beidellite series, with very little vermiculite forming the oxisol-vertisol horizon.

Reduction of structural Fe in Na-exchanged dioctahedral smectites decreases swellability in water, but because clay interlayers also collapse in the process the concomitant effect on surface hydration energy is uncertain. This study examined the hydration behavior of oxidized and reduced dioctahedral smectite clays exchanged with polar (Na) and weakly-polar (organic) cations to determine the nature of the surface before and after Fe reduction, and to determine if clay surfaces are hydrophilic or hydrophobic. The H2O content in various dioctahedral smectites decreased if Na was replaced by tetramethylammonium (TMA), trimethylphenylammonium (TMPA), or hexadecyltrimethylammonium (HDTMA). Among the organo-clays, H2O adsorption decreased with increasing complexity of the cation. For oxidized smectites, those exchanged with TMPA retained less H2O than those exchanged with Na at all pressures. The extent of this difference depended on the clay and decreased with increasing applied pressure. Reduction of Fe(III) to Fe(II) in the octahedral sheets decreased the swelling of Na-saturated smectites, apparently causing some previously swelling interlayers to collapse. If the Na interlayer cation was exchanged to alkylammonium after reduction, but prior to swelling-pressure measurements, the swelling increased or remained near constant, suggesting that the organo-cation disrupted the collapse process of the interlayers associated with the reduced smectite layers. Reduced TMPA-saturated smectite surfaces are more strongly hydrated if the octahedral sheet is reduced than if oxidized. Thus, reduction of structural Fe increases the hydration energy of smectite basal surfaces, but swellability could decrease or increase depending on the extent of interlayer collapse occurring with different exchangeable cations.

Low-temperature chlorites formed in diagenetic to low-grade metamorphic environments generally have greater Si contents and larger numbers of octahedral vacancies, and smaller Fe+Mg contents than higher-grade metamorphic chlorites. The compositional variations are characterized approximately by four end-member components: Al-free trioctahedral chlorite, chamosite, corundophilite, and sudoite. The solid solution is considered to be a random mix of cations and vacancies in the octahedral sites. Using the compositions of chlorites from Niger, Rouez, and Saint Martin diagenetic-hydrothermal series, a new, more convenient geothermometer, applicable to low-T chlorites is proposed and comparison made with geothermometers proposed previously. The chlorites studied contain appreciable amounts of Fe(III) (>14% of the total Fe), determined by Mössbauer spectroscopy. The calculations under which all Fe was regarded as ferrous gave considerable overestimates for the formation temperature, irrespective of the geothermometer used. This problem was reduced by taking into account the presence of Fe(III) in the octahedral sites. The geothermometer from this study gave more reasonable estimates than the geothermometers proposed by Walshe (1986) and Vidal et al. (2001), particularly in the case of the Niger chlorites which crystallized in the lowest-temperature conditions. The ordered-site substitution model of solid solution developed by Vidal et al. (2001) predicted satisfactorily the formation temperature of the Rouez chlorites and of some of the Saint Martin chlorites, suggesting that the chlorite compositions are controlled by the exchange at low-T conditions while they are controlled by Tschermak exchange at higher temperatures. The decreasing number of vacancies with temperature are poorer in Fe-rich than in Fe-poor chlorites. Furthermore, the ordered-site occupation of cations and vacancies in trioctahedral chlorite occurs concomitantly with the compositional changes ruled by increasing temperature conditions.

We investigated the compositions of a suite of 361 chrome-bearing spinels from spinel peridotites, ophiolitic mantle chromitites and from layered igneous intrusions in which the Fe2+/Fe3+ ratio has been determined by Mössbauer spectroscopy. We explore the crystal-chemical controls on the distribution of Fe3+ and on mineral stoichiometry with regard to electron-probe correction procedures for estimating Fe3+ in spinel. We find that chrome-bearing spinels can be subdivided into three groups: a Cr–Al-rich group; a high-Fe group; and a highly oxidised group. Spinels of the Cr–Al group are found in spinel peridotites and ophiolitic mantle chromitites. They are normal spinels with a low Fe3+ content and compositions that are close to stoichiometric. Spinels in the high-Fe group are found in layered igneous intrusions. They are also normal spinels which have higher concentrations of Fe3+ on the octahedrally coordinated site than is found for the Cr–Al group. Stoichiometric calculations tend to overestimate the Fe3+ content and their compositions do not conform to the MgO–cr# correlation found in the Cr–Al group. Spinels in the highly oxidised group are found in layered intrusions and ophiolitic mantle chromitites. They have (Fe3+/ΣFe)Möss > 0.4 and high Cr (cr# > 0.5), but relatively low Fe2+ (fe# 0.24–0.56). Stoichiometric calculations tend to underestimate the Fe3+ content. They represent normal spinels in which tetrahedrally coordinated Fe2+ has been oxidised to Fe3+.

Our data show that spinels with greater Cr and Fe are sufficiently different in their crystal chemistry from the aluminous spinels to indicate that the EPMA correction procedures developed for Fe3+ in aluminous spinels on the basis of the Cr/Al ratio, and used in oxy-thermobarometry, are inappropriate for Cr-rich and Fe-rich compositions.

Perovskite compositions are used to investigate the relationship between the minor components (i.e. LREE, Fe3+ and Nb) and the oxygen fugacity (fo2) of perovskite in four different kimberlite lithofacies from the Dutoitspan pipe, Kimberley, South Africa, which range from diamondiferous to barren. The perovskite textures and chemical variations provide insight into magmatic and eruptive processes. Some crystals display cores with rims separated by a sharp boundary. The cores contain larger Na and LREE contents relative to the rims, which show a large increase in Fe3+ and Al. The mid-grade and barren kimberlites have bi-modal cores, reflected in the mineral chemistry, signifying multiple batches of magma and magma mixing. The fo2 of the magma is determined by an Fe-Nb oxygen barometer. The most diamondiferous kimberlite has the greatest Fe3+ content and highest fo2 (NNO –3.6 to –1.1). The kimberlite containing large diamonds has the smallest Fe3+ content and lowest fo2 (NNO –5.2 to –3.0). The barren and mid-grade kimberlites display a wide range of fo2,(NNO –5.3 to –1.5) as a result of perovskites forming in different melts and subsequently mixing together. Chemical and petrological evidence suggests that the volatile content, degassing, decompression and rate of crystallization can influence the rate at which the magma is erupted. One possibility is that the most oxidized magma, containing the highest volatile content, is therefore erupted much more rapidly, preserving diamond as a consequence.

The ferromagnesian silicate minerals, such as garnets, pyroxenes, micas, and amphiboles, appear in a variety of geothermometers and geobarometers. Where complete chemical analyses are available and regardless of bulk composition (metamorphosed pelitic or mafic), the aforementioned minerals commonly contain ferric iron. In mineral analyses using the electron microprobe, ferric and ferrous iron are not distinguished, and all the iron is treated as FeO. In ferric Fe-bearing minerals, this treatment results in (1) low analytical sums and (2) excess cations in the mineral formulae. Assuming ideal stoichiometry (ideal formula cations and oxygens) allows direct ferric estimates in garnets and pyroxenes; amphiboles require additional assumptions concerning site occupancies, and, for micas, no acceptable constraint exists for a ferric estimate. Based on ferric iron determinations for some metamorphic ferromagnesian silicates, the proportion of ferric to total iron increases at higher XMg values. The influence of ferric estimates on T and P calculations depends on the model used and on the extent the ferric estimate alters the relative proportions of end-members. Several examples suggest that, in general, if ferric estimates (or determinations) are made, they should be made for all the relevant minerals.

Titanomaghemite occurs in a relatively fresh doleritic intrusion in an area of Precambrian gneiss in Minas Gerais, Brazil. It hosts exsolution lamellae of ilmenite and contains more than 90% of the iron in the ferric form. It is more resistant to weathering than the ilmenite and is inherited virtually unaltered by the resulting soils. Titanomaghemite, extracted as grains from a weathered rind of the rock, has lattice parameter a0 = 0.8348(3) nm and has a canted spin structure due to substitution of non-magnetic ions on tetrahedral and octahedral sites of the spinel structure. The average canting angle is 32 ± 3° and canting occurs predominantly on the octahedral iron sublattice. Its formula, based on microprobe analysis and Mössbauer spectroscopy may be expressed as:

where [] and {} denote ions on tetrahedral and octahedral sites, respectively. The spontaneous magnetization of the mineral is 36(3) J/T/kg.